A system, apparatus and method of analyzing blood material

Abstract

A system and method for analyzing blood material may include controlling a first pumping mechanism to draw blood material, associated with a specific subject, from a blood tube into a microfluidic chip; acquiring, via an imaging device, at least one image depicting cells in the microfluidic chip; based on the at least one image, controlling a Raman spectroscopy device to obtain one or more Raman spectra from a cell of interest, pertaining to a predetermined cell type; and analyzing the one or more Raman spectra, to determine a subtype of the at least one cell of interest.

Description

A SYSTEM, APPARATUS AND METHOD OF ANALYZING BLOOD MATERIALCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority of U.S. Patent Application No. 63 / 660,584, filedJune 17, 2024, which is hereby incorporated by reference in its entirety.FIELD OF THE INVENTION

[0002] The present invention relates generally to automated medical devices. More specifically, the present invention relates to a device, system and method of analyzing blood material.BACKGROUND

[0003] Currently, advanced blood cell phenotyping relies on Flow Cytometry, which requires a specialized lab, expert team, manual workflows, and expensive reagents, making it impractical for routine hospital use.SUMMARY OF THE INVENTION

[0004] Embodiments of the invention may introduce novel clinical screening capabilities in-line with Complete Blood Count (CBC) testing. As elaborated herein, embodiments of the invention may integrate Al-driven automated spectroscopy, to simplify the process of clinical screening, rendering the need for flow-cytometry redundant, and adding previously unavailable data to routine CBC results.

[0005] Embodiments of the invention may facilitate a range of CBC-reflex tests, including for example the analysis of lymphocyte T-cells, B-cells, and Natural Killer (NK) cells (TBNK) subsets, detection of hemoglobin abnormalities, and quantification of the immune system activation profile. These tests may be integrated with CBC results to effectively screen for clinical conditions such as hematological malignancies and Sepsis.

[0006] According to some embodiments, at least one processor may be configured to control a first pumping mechanism to draw blood material, associated with a specific subject, from a blood tube into a microfluidic chip. The processor may be further configured to acquire, via an imaging device, at least one image depicting cells in the microfluidic chip. In some embodiments, the processor may also be adapted to, based on the at least one image, control aRaman spectroscopy device to obtain one or more Raman spectra from a cell of interest, pertaining to a predetermined cell type. The processor may subsequently analyze the one or more Raman spectra, to determine a subtype of the at least one cell of interest.

[0007] According to some embodiments, the processor may apply an object recognition algorithm on the at least one image, to identify a plurality of cells of interest, as pertaining to the predetermined cell type. Based on the identification, the processor may calculate a transition pattern of the Raman spectroscopy device. The processor may control one or more actuators, to adjust a position of the Raman spectroscopy device based on the transition pattern. The processor may also be adapted to control the Raman spectroscopy device to specifically illuminate one or more cells of the plurality of cells of interest, so as to obtain the one or more Raman spectra from each of the one or more cells of interest.

[0008] According to some embodiments, the processor may be configured to receive a dataset including a plurality of Raman spectra associated with a respective plurality of cells. The processor may further receive a plurality of subtype annotations, respectively associated with the plurality of cells, where each subtype annotation may indicate a specific cell subtype. In some embodiments, the processor may be configured to use the one or more subtype annotations as supervisory information, to train a Machine Learning (ML)-based, subtype classification model, to predict a subtype of an incident cell, based on the received dataset.

[0009] According to some embodiments, the processor may be configured to receive one or more Raman spectra of a target cell. The processor may provide the one or more Raman spectra as input to the subtype classification model. The processor may subsequently obtain, from the subtype classification model, a subtype of the target cell, based on said training.

[0010] According to some embodiments, the processor may be configured to receive a dataset that may include: (a) a plurality of Raman spectra associated with a respective plurality of cells, and (b) a plurality of images, respectively depicting the plurality of cells. The processor may receive a plurality of subtype annotations, respectively associated with the plurality of cells, where each annotation may indicate a specific cell subtype. In some embodiments, the processor may use the one or more subtype annotations as supervisory information, to train an ML-based, subtype classification model to predict a subtype of an incident cell, based on the received dataset.

[0011] According to some embodiments, the processor may be configured to receive one or more Raman spectra of a target cell. The processor may provide the one or more Raman spectra as input to the subtype classification model. In some embodiments, the processor may further receive, via the imaging device, an image depicting the target cell. The processor may provide said image of the target cell as input to the subtype classification model. The processor may further obtain, from the subtype classification model, a subtype of the target cell of interest, based on said training.

[0012] According to some embodiments, the processor may be configured to, for each image of the plurality of images, apply an image analysis algorithm, to obtain at least one morphological feature of the respective depicted cell. The processor may use the one or more subtype annotations as supervisory information, to train an ML-based, subtype classification model to predict a subtype of an incident cell, based on (a) the plurality of Raman spectra of the dataset, and (b) the at least one morphological feature of the plurality of images.

[0013] Additionally, or alternatively, the processor may be configured to receive one or more Raman spectra of a target cell, and provide the one or more Raman spectra as input to the subtype classification model. The processor may be configured to receive, via the imaging device, an image depicting the target cell, and may apply the image analysis algorithm on said image, to obtain at least one morphological feature of the depicted target cell. The processor may provide the at least one morphological feature as input to the subtype classification model, and may obtain, from the subtype classification model, a subtype of the target cell, further based on the at least one morphological feature, in accordance with said training.

[0014] According to some embodiments, the processor may determine a subtype of a plurality of cells in the blood material of the subject. The processor may be further adapted to, for one or more (e.g., each) cell subtype, compute a subtype percentage or proportion, representing percentage of that subtype among cells of the same type (or other cell types in the depicted sample).

[0015] In some embodiments, the processor may, for each cell subtype, evaluate a subtype quantity value, representing a number of cells of that subtype in the at least one image. The processor may also be adapted to determine a condition of the subject based on (i) said subtype percentage and (ii) said subtype quantity value.

[0016] According to some embodiments, the processor may be configured to receive, from a Complete Blood Count (CBC) module, at least one CBC report associated with the subject. The processor may further provide the at least one CBC report as a first input to a pretrained, ML- based, subject classification model. The processor may provide at least one of (i) a subtype percentage of at least one subtype and (ii) a subtype quantity value of at least one subtype, as a second input to the subject classification model.

[0017] The processor may also be adapted to obtain, from the subject classification model, a prediction of the condition of the subject, based on the first and second input.

[0018] According to some embodiments, the processor may be configured to receive a second dataset including (i) a plurality subtype percentages, associated with a cohort of subjects, (ii) a plurality of subtype quantity values associated with the cohort of subjects, and (iii) a plurality of CBC report data elements associated with the cohort of subjects. The processor may be further adapted to receive a plurality of condition annotations, respectively associated with each of the cohort of subjects, wherein each condition annotation may indicate a medical condition of the respective subject. In some embodiments, the processor may be configured to use the plurality of condition annotations, to train the ML-based subject classification model to predict the medical condition of an incident subject, based on the second dataset.

[0019] According to some embodiments, the processor may be configured to obtain a predetermined number of Raman spectra, corresponding to a predetermined number of cells in the blood material. The processor may control a second pumping mechanism to flush a cleaning solution via the microfluidic chip, thereby preparing the microfluidic chip to receive blood material from another blood tube.

[0020] According to some embodiments, the first pumping mechanism may include a puncturing device, a first pump and an inlet valve. The processor may be configured to control the puncturing device, to enter the blood tube. The processor may control the first pump, to draw blood material from the blood tube via the puncturing device. In some embodiments, the processor may be configured to open the inlet valve, to insert a sample of the blood material into the microfluidic chip. The processor may also be adapted to close the inlet valve, to stabilize the sample of blood material in the microfluidic chip.

[0021] According to some embodiments, the processor may be configured to control the imaging device, so as to auto-focus on the stabilized sample of blood, in a segment of themicrofluidic chip. The processor may be further adapted to acquire at least one first image depicting cells in the segment of the microfluidic chip.

[0022] According to some embodiments, the processor may be configured to obtain, from an automated CBC module, at least one CBC report selected from (i) a hematology analysis report and (ii) a blood-chemistry analysis report. The processor may be further adapted to identify at least one anomaly in the subject's blood material based on the CBC report data element. In some embodiments, the processor may be configured to, based on the identification of said at least one anomaly, control at least one of the Raman spectroscopy device and first pumping device, to obtain the Raman spectrum.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The subject matter regarded as the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. The invention, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings in which:

[0024] Fig. 1 is a block diagram, depicting a computing device which may be included in a system for analyzing blood material, according to some embodiments;

[0025] Fig. 2 is a block diagram, depicting a system for analyzing blood material, according to some embodiments;

[0026] Fig. 3 is a flow diagram, depicting a method of analyzing blood material, according to some embodiments;

[0027] Fig. 4 is a flow diagram, depicting an example of analysis of blood material by the system for analyzing blood material, according to some embodiments of the invention;

[0028] Fig. 5 is a flow diagram, depicting an example of an acquisition process, for analyzing blood material by embodiments of the system according to some embodiments of the invention; and

[0029] Fig. 6 is a flow diagram, depicting an example of a cleaning process for a microfluidic subsystem of the system for analyzing blood material, according to some embodiments of the invention.

[0030] It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elementsmay be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.DETAILED DESCRIPTION OF THE PRESENT INVENTION

[0031] One skilled in the art will realize the invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The foregoing embodiments are therefore to be considered in all respects illustrative rather than limiting of the invention described herein. Scope of the invention is thus indicated by the appended claims, rather than by the foregoing description, and all changes that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

[0032] In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be understood by those skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the present invention. Some features or elements described with respect to one embodiment may be combined with features or elements described with respect to other embodiments. For the sake of clarity, discussion of same or similar features or elements may not be repeated.

[0033] Although embodiments of the invention are not limited in this regard, discussions utilizing terms such as, for example, “processing,” “computing,” “calculating,” “determining,” “establishing”, “analyzing”, “checking”, or the like, may refer to operation(s) and / or process(es) of a computer, a computing platform, a computing system, or other electronic computing device, that manipulates and / or transforms data represented as physical (e.g., electronic) quantities within the computer’s registers and / or memories into other data similarly represented as physical quantities within the computer’s registers and / or memories or other information non-transitory storage medium that may store instructions to perform operations and / or processes.

[0034] Although embodiments of the invention are not limited in this regard, the terms “plurality” and “a plurality” as used herein may include, for example, “multiple” or “two or more”. The terms “plurality” or “a plurality” may be used throughout the specification to describe two or more components, devices, elements, units, parameters, or the like. The term “set” when used herein may include one or more items.

[0035] Unless explicitly stated, the method embodiments described herein are not constrained to a particular order or sequence. Additionally, some of the described method embodiments or elements thereof can occur or be performed simultaneously, at the same point in time, or concurrently.

[0036] Reference is now made to Fig. 1, which is a block diagram depicting a computing device, which may be included within an embodiment of a system for analyzing blood material, according to some embodiments.

[0037] Computing device 1 may include a processor or controller 2 that may be, for example, a central processing unit (CPU) processor, a chip or any suitable computing or computational device, an operating system 3, a memory 4, executable code 5, a storage system 6, input devices 7 and output devices 8. Processor 2 (or one or more controllers or processors, possibly across multiple units or devices) may be configured to carry out methods described herein, and / or to execute or act as the various modules, units, etc. More than one computing device 1 may be included in, and one or more computing devices 1 may act as the components of, a system according to embodiments of the invention.

[0038] Operating system 3 may be or may include any code segment (e.g., one similar to executable code 5 described herein) designed and / or configured to perform tasks involving coordination, scheduling, arbitration, supervising, controlling or otherwise managing operation of computing device 1, for example, scheduling execution of software programs or tasks or enabling software programs or other modules or units to communicate. Operating system 3 may be a commercial operating system. It will be noted that an operating system 3 may be an optional component, e.g., in some embodiments, a system may include a computing device that does not require or include an operating system 3.

[0039] Memory 4 may be or may include, for example, a Random-Access Memory (RAM), a read only memory (ROM), a Dynamic RAM (DRAM), a Synchronous DRAM (SD-RAM), a double data rate (DDR) memory chip, a Flash memory, a volatile memory, a non-volatile memory, a cache memory, a buffer, a short term memory unit, a long term memory unit, or other suitable memory units or storage units. Memory 4 may be or may include a plurality of possibly different memory units. Memory 4 may be a computer or processor non-transitory readable medium, or a computer non- transitory storage medium, e.g., a RAM. In one embodiment, a non-transitory storage medium such as memory 4, a hard disk drive, another storage device, etc. may store instructions or code which when executed by a processor may cause the processor to carry out methods as described herein.

[0040] Executable code 5 may be any executable code, e.g., an application, a program, a process, task, or script. Executable code 5 may be executed by processor or controller 2 possibly under control of operating system 3. For example, executable code 5 may be an application that may analyze blood material as further described herein. Although, for the sake of clarity, a single item of executable code 5 is shown in Fig. 1, a system according to some embodiments of the invention may include a plurality of executable code segments similar to executable code 5 that may be loaded into memory 4 and cause processor 2 to carry out the methods described herein.

[0041] Storage system 6 may be or may include, for example, a flash memory as known in the art, a memory that is internal to, or embedded in, a micro controller or chip as known in the art, a hard disk drive, a CD-Recordable (CD-R) drive, a Blu-ray disk (BD), a universal serial bus (USB) device or other suitable removable and / or fixed storage unit. Data pertaining to blood material may be stored in storage system 6 and may be loaded from storage system 6 into memory 4 where it may be processed by processor or controller 2. In some embodiments, some of the components shown in Fig. 1 may be omitted. For example, memory 4 may be a non-volatile memory having the storage capacity of storage system 6. Accordingly, although shown as a separate component, storage system 6 may be embedded or included in memory 4.

[0042] Input devices 7 may be or may include any suitable input devices, components, or systems, e.g., a detachable keyboard or keypad, a mouse and the like. Output devices 8 may include one or more (possibly detachable) displays or monitors, speakers and / or any other suitable output devices. Any applicable input / output (VO) devices may be connected to Computing device 1 as shown by blocks 7 and 8. For example, a wired or wireless network interface card (NIC), a universal serial bus (USB) device or external hard drive may be included in input devices 7 and / or output devices 8. It will be recognized that any suitable number of input devices 7 and output device 8 may be operatively connected to Computing device 1 as shown by blocks 7 and 8.

[0043] A system according to some embodiments of the invention may include components such as, but not limited to, a plurality of central processing units (CPU) or any other suitable multi-purpose or specific processors or controllers (e.g., similar to element 2), a plurality of input units, a plurality of output units, a plurality of memory units, and a plurality of storage units.

[0044] The term neural network (NN) or artificial neural network (ANN), e.g., a neural network implementing a machine learning (ML) or artificial intelligence (Al) function, may be used herein to refer to an information processing paradigm that may include nodes, referred to as neurons, organizedinto layers, with links between the neurons. The links may transfer signals between neurons and may be associated with weights. A NN may be configured or trained for a specific task, e.g., pattern recognition or classification. Training a NN for the specific task may involve adjusting these weights based on examples. Each neuron of an intermediate or last layer may receive an input signal, e.g., a weighted sum of output signals from other neurons, and may process the input signal using a linear or nonlinear function (e.g., an activation function). The results of the input and intermediate layers may be transferred to other neurons and the results of the output layer may be provided as the output of the NN. Typically, the neurons and links within a NN are represented by mathematical constructs, such as activation functions and matrices of data elements and weights. At least one processor (e.g., processor 2 of Fig. 1) such as one or more CPUs or graphics processing units (GPUs), or a dedicated hardware device may perform the relevant calculations.

[0045] Reference is also made to Fig. 2, which is a block diagram depicting a system 10 for analyzing blood material, according to some embodiments of the invention.

[0046] According to some embodiments of the invention, system 10 may be implemented as a software module, a hardware module, or any combination thereof. For example, system 10 may be, or may include a computing device such as element 1 of Fig. 1, and may be adapted to execute one or more modules of executable code (e.g., element 5 of Fig. 1) to analyze blood material, as further described herein.

[0047] As shown in Fig. 2, arrows may represent flow of one or more data elements to and from system 10 and / or among modules or elements of system 10. Additionally, some arrows of Fig. 2 may represent physical interrelations between modules or elements of system 10. Some arrows have been omitted on Fig. 2 for the purpose of clarity.

[0048] As shown in Fig. 2, system 10 may implement a comprehensive workflow, or line, which integrates between different modules, disciplines, and technologies. Such integration between Complete Blood Cell (CBC) modules, and Raman-based technology for analyzing blood material manifests a synergy that improves inline analysis of blood samples, allowing improvement in accuracy, cost-effectiveness and near-immediacy of automated blood sample analyses.

[0049] In some embodiments, system 10 may integrated as a unified apparatus, that may integrate functionality of several disciplines, and / or subsystems, facilitating the benefits of the present invention in a single location. These subsystems may include (i) a sampling subsystem 50, configured to sample blood material from blood sampling component, (ii) a CBC module 30, configured toproduce a CBC report elaborating components of the sampled blood material, (iii) a microfluidics chip 150, adapted to maintain the sampled blood material for analysis (iv) at least one imaging device 160, configured to produce microscopy image(s) 160M of the sampled blood material in chip 150, (v) at least one Raman spectroscopy device 140, adapted to produce Raman spectra of specific locations (e.g., specific cells) in the chip 150, e.g., based on analysis of image 160M, and at least one processor or controller (used herein interchangeably, according to context) 2, adapted to analyze CBC, Raman spectra, images 160M and / or Laboratory Information System (LIS) information 40, to classify the sampled blood material according to specific use cases.

[0050] As shown in Fig. 2, system 10 may include a first (“inlet”) pumping mechanism 110, a microfluidic chip 150, a microscopic imaging device 160, and a Raman spectroscopy device 140.

[0051] Microfluidic chip 150 may be, or may include for example, a Calcium Fluoride chip, configured as a closed 2-dimensional channel.

[0052] Microscopic imaging device 160 may be, or may include for example, a bright field microscope, a dark field microscope, a phase contrast microscope, a DIC microscope, a DPC microscope, and the like.

[0053] Additionally, system 10 may include a non-transitory memory device (e.g., memory 4 of Fig. 1), where modules of instruction code are stored, and at least one processor (e.g., processor of Fig. 1). The at least one processor 2 may be configured to execute the modules of instruction code to analyze blood material 25 as described herein.

[0054] Blood material 25 may include blood as sampled from a subject (human or animal), or any substance or compound extracted therefrom. For example, blood material 25 may include any compound that includes White Blood Cells (WBC), Red Blood Cells (RBC), plasma, and the like.

[0055] According to some embodiments, processor 2 may control the first pumping mechanism to draw blood material, associated with a specific subject, from a blood tube 20 into microfluidic chip 150.

[0056] For example, inlet pumping mechanism 110 may include a puncturing device 110PD, a first pump 110PU and an inlet valvel 10V. Processor 2 may control the puncturing device, to enter blood tube 20, and control the first pump, to draw blood material from blood tube 20 via the puncturing device. Processor 2 may then open the inlet valve, to insert a sample of the blood material into microfluidic chip 150, and close the inlet valve back, to stabilize the sample of blood material in the microfluidic chip.

[0057] Processor 2 may then control imaging device 160 to acquire at least one image 160M, depicting cells in microfluidic chip 150. For example, processor 2 may control imaging device 160 so as to auto-focus on the stabilized sample of blood, in a segment of the microfluidic chip 150, and acquire at least one image 160M depicting cells in the segment of microfluidic chip 150.

[0058] As elaborated herein, processor 2 may control Raman spectroscopy device 140 based on the at least one image 160M, to obtain one or more Raman spectra 140SP from cells of interest, pertaining to a predetermined cell type.

[0059] Processor 2 may subsequently analyze the one or more Raman spectra 140SP, to determine a subtype 170ST of the at least one cell of interest. For example, processor 2 may apply a Machine Learning (ML)-based, subtype classification model 170 on data originating from, or representing Raman spectra 140SP, to determine, or predict subtype 170ST of the at least one cell of interest.

[0060] As used herein, a "subtype" of cells may refer to a distinct category or classification within a broader cell type, characterized by specific features, functions, or markers. Subtypes may be officially recognized in the art, such as lymphocytes, neutrophils, and monocytes being subtypes of white blood cells. Additionally, subtypes as referred to in this invention may include classifications that are not yet officially recognized but are defined by a combination of morphological features and spectroscopic characteristics. For example, a subtype may be defined by specific morphological features 185 identified by the image analyzer 180, combined with unique Raman spectral signatures 140SP obtained by the Raman spectroscopy device 140. These novel subtypes may represent cells in particular functional states, activation levels, or exhibiting specific molecular characteristics that are relevant for diagnostic or research purposes. The subtype classifier 170 may use these combined features to categorize cells into subtypes that provide more granular and potentially clinically relevant information beyond traditional cell type classifications.

[0061] According to some embodiments, processor 2 may apply an image analysis algorithm, such as an object recognition algorithm 182 on the at least one image 160M. Processor 2 may thereby identify a plurality of cells of interest, as pertaining to specific, one or more predetermined cell types 182T.

[0062] Based on the identification of cell type 182T, processor 2 may calculate a transition pattern 130T of the Raman spectroscopy device 140. For example, transition pattern 130T may be a data element representing a course, or path, by which Raman spectroscopy device 140 may efficientlytraverse through a number of locations on microfluidic chip 150, where cells of interest of the predetermined type 182T are located.

[0063] Processor 2 may subsequently control one or more actuators 130 to adjust a position of the Raman spectroscopy device 140 based on the transition pattern 130T, e.g., so as to follow the locations of the cells of interest.

[0064] According to some embodiments, processor 2 may subsequently control the Raman spectroscopy device 140 to specifically illuminate one or more (e.g., a single) cells of the plurality of cells of interest, so as to uniquely obtain the one or more Raman spectra from each of the one or more cells of interest.

[0065] According to some embodiments, processor 2 may use a supervised training scheme to train subtype classifier 170, based on an annotated training dataset.

[0066] For example, processor 2 may (e.g. , during a training stage) receive a dataset that includes a plurality of Raman spectra associated with a respective plurality of cells. Processor 2 may also receive a plurality of subtype annotations 170AN, respectively associated with the plurality of cells. Each subtype annotation may indicate a specific cell subtype. Processor 2 may use the one or more subtype annotations 170AN as supervisory information, to train ML-based, subtype classification model 170 so as to predict a subtype 170ST of an incident cell, based on the received dataset.

[0067] Processor 2 may (e.g., in a subsequent inference stage) receive one or more Raman spectra 140SP pertaining to a target cell of interest. Processor 2 may provide the one or more Raman spectra 140SP as input to the subtype classification model, and obtain, from the subtype classification model 170, a subtype 170ST of the target cell, based on said training.

[0068] For example, object recognition algorithm may identify a cell as pertaining to a T-Cell type 182T, and subtype classification model 170 may be applied on Raman spectra 140SP acquired from that cell, to determine the cell subtype 170ST as CD3+, CD4+, or CD8+.

[0069] In another example, object recognition algorithm may identify a cell as pertaining to a B- Cell type 182T, and subtype classification model 170 may be applied on Raman spectra 140SP acquired from that cell, to determine the cell subtype 170ST as CD3- or CD19+.

[0070] In yet another example, object recognition algorithm may identify a cell as pertaining to an NK-Cell type 182T, and subtype classification model 170 may be applied on Raman spectra 140SP acquired from that cell, to determine the cell subtype 170ST as CD3- or CD56+.

[0071] Additionally, or alternatively, subtype classification model 170 may be trained, and applied further based on cell images 160M.

[0072] In other words, processor 2 may receive a training dataset comprising: (a) a plurality of Raman spectra 140SP associated with a respective plurality of cells, and (b) a plurality of images 160M, respectively depicting the plurality of cells. Processor 2 may also receive a plurality of subtype annotations 170AN, respectively associated with the plurality of cells, wherein each annotation 170AN indicates a specific cell subtype. Processor 2 may use the one or more subtype annotations 170AN as supervisory information, to train ML-based, subtype classification model 170, so as to predict subtype 170ST of incident cell(s), based on the received training dataset.

[0073] Processor 2 may (e.g., in a subsequent inference stage) receive one or more Raman spectra 140SP pertaining to a target cell of interest. Processor 2 may provide the one or more Raman spectra 140SP as a first input to subtype classification model 170. Processor 2 may also receive, via imaging device 160, an image depicting the target cell 160M, and provide image 160M of the target cell as a second input to the subtype classification model 170. Processor 2 may thereby obtain from subtype classification model 170, a subtype 170ST of the target cell of interest, based on the training of ML model 170.

[0074] Additionally, or alternatively, subtype classification model 170 may be trained, and applied further based on morphological features extracted from cell images 160M.

[0075] For example, processor 2 may (e.g., during a training stage), apply an image analysis algorithm 180 on one or more (e.g., each) image 160M of the plurality of images 160M, to obtain at least one morphological feature 185 pertaining to the respective depicted cell. Processor 2 may use the one or more subtype annotations 170AN as supervisory information, to train ML-based, subtype classification model 170 to predict subtype 170ST of an incident cell, based on (a) the plurality of Raman spectra 140SP of the training dataset, and (b) the at least one morphological feature 185 extracted from each of the plurality of images 160M.

[0076] Processor 2 may (e.g., in a subsequent inference stage) receive one or more Raman spectra 140SP of a target cell, and provide the one or more Raman spectra 140SP as input to subtype classification model 170. Processor 2 may also receive, via imaging device 160, an image depicting the target cell, and apply image analysis algorithm 180 on image 160M, to obtain at least one morphological feature 185 of the depicted target cell. Processor 2 may then provide the at least one morphological feature 185 as a second input to the subtype classification model. Processor 2 maythereby obtain from subtype classification model 170 a subtype 170ST of the target cell, based on training of ML model 170.

[0077] According to some embodiments, system 10 may determine a condition of the subject based on the determination of cell subtypes. For example, processor 2 may determine a subtype of a plurality of cells in the blood material of the subject, as elaborated herein. For each cell subtype 170ST, processor 2 may compute a subtype percentage value 170PC, representing percentage of that subtype 170ST among cells of the same type 182T.

[0078] Additionally, or alternatively, for each cell subtype 170ST, processor 2 may evaluate a subtype quantity value 170QU, representing a number of cells of that subtype in the at least one image 160M, which may also represent a quantity of cells of that subtype in tube 20.

[0079] As elaborated herein, processor 2 may subsequently determine a medical condition 190SC of the subject based on (i) said subtype percentage 170PC and (ii) said quantity value 170QU.

[0080] Additionally, or alternatively, processor 2 may determine a medical condition 190SC of the subject based on (i) said subtype percentage 170PC and (ii) said quantity value 170QU, and / or further based on data acquired from a Complete Blood Count (CBC) report.

[0081] As shown in Fig. 2, system 10 may be associated with a CBC module 30. Additionally, or alternatively, system 10 may include CBC module 30, such that processor 2 may control CBC module 30, to obtain therefrom a CBC report data element 3 OR, as known in the art.

[0082] According to some embodiments, processor 2 may receive from CBC module 30, at least one CBC report 3 OR associated with the subject. Processor 2 may provide the at least one CBC report 3 OR as a first input to a pretrained, second ML-based model, denoted herein as “subject classification model 190”.

[0083] Processor 2 may further provide at least one of (i) a subtype percentage of at least one subtype and (ii) a subtype quantity value of at least one subtype, as a second input to the subject classification model 190. Processor 2 may thereby obtain from subject classification model 190, a prediction of the medical condition 190SC of the subject, based on the first and second inputs.

[0084] According to some embodiments, processor 2 may use a supervised training scheme to train subject classification model 190, based on a second, annotated training dataset.

[0085] For example, processor 2 may (e.g., during a training stage) receive a second training dataset, that may include (i) a plurality subtype percentages 170PC, associated with a cohort ofsubjects, (ii) a plurality of subtype quantity values 170QU, associated with the cohort of subjects, and / or (iii) a plurality of CBC report data elements 30R, associated with the cohort of subjects.

[0086] Processor 2 may further receive a plurality of condition annotations 190AN, respectively associated with each of the cohort of subjects. Each condition annotation 190AN may indicate a medical condition of the respective subject.

[0087] Processor 2 may subsequently use the plurality of condition annotations 190AN, to train ML-based subject classification model 190 to predict the medical condition 190SC of an incident subject, based on the second training dataset.

[0088] As elaborated herein, system 10 may allow automated blood sample analyses by CBC and / or Raman spectroscopy. System 10 may therefore provide means for streamlining laboratory workflow, vis-a-vis standard blood sample tubes.

[0089] For example, system 10 may include a second (“outlet”) pumping mechanism 120, that may include a second pump 120PU and a second valve 120V. Processor 2 may obtain a predetermined number of Raman spectra 140SP, corresponding to a predetermined number of cells in the blood material. Having counted the number of required Raman spectra data elements 140SP, processor 2 may configure system 10 to proceed to a next blood sample. Processor 2 may therefore control the outlet pumping mechanism 120 to flush a cleaning solution via microfluidic chip 150, thereby preparing the microfluidic chip to receive blood material from another, subsequent blood tube 20.

[0090] As elaborated herein, system 10 may include, or may be communicatively connected or associated with an automated CBC module 30. Processor 2 may obtain from CBC module 30, at least one CBC report 3 OR which may include a hematology analysis report, a blood-chemistry analysis report, and the like. According to some embodiments, system 10 may use CBC report 30Rto screen, for suspicious anomalies, and employ Raman spectroscopy only on selected, or suspicious samples.

[0091] In other words, processor 20 may identify at least one anomaly in the subject’s blood material based on the CBC report 20R data element. Based on the identification of such an anomaly, processor 20 may apply a reflex rule, as commonly referred to in the art, to control at least one of inlet pumping device 110 and Raman spectroscopy device 140 to sample blood material and obtain therefrom Raman spectra 140SP for analysis.

[0092] For example, CBC report 20R data element may indicate a condition of lymphocytosis (e.g., a state of high lymphocyte cell count). System 10 may thus apply a reflex rule to trigger analysis of lymphocyte type 182T cells by Raman 140 technology. Processor 2 may therefore calculate a ratiobetween T subtype 170ST and B subtype 170ST lymphocytes 182T, to determine a medical condition 190SC of the subject.

[0093] The following are non-limiting examples of classification of subject medical condition, based on a synergic combination of CBC report data 3 OCR, and Raman spectroscopy based, cell subtype analysis.

[0094] Embodiments of system 10 may detect Iron Deficiency Anemia based on CBC report data 3 OR that includes MCV, hemoglobin concentration (MCHC), and Platelet count, alongside Raman spectroscopy based, cell subtype analysis of RBC cell profiles.

[0095] In another example, embodiments of system 10 may detect MDS (Pre-cancerous state) based on CBC report data 30R that includes a 5-class count, alongside Raman spectroscopy based, cell subtype analysis of Neutrophil activation state.

[0096] In another example, embodiments of system 10 may provide early detection of high-risk Sepsis based on CBC report data 30R that includes Neutrophil / Lymphocyte ratio alongside Raman spectroscopy based, cell subtype analysis of an immune system activation state.

[0097] In yet another example, embodiments of system 10 may provide CLL flagging based on CBC report data 3 OR that includes a Lymphocyte count, alongside Raman spectroscopy based, cell subtype analysis of B cells profile.

[0098] Reference is now made to Pig. 3, which is a flow diagram, depicting a method of analyzing blood material by at least one processor, according to some embodiments of the invention.

[0099] As shown in step S1005, the at least one processor (e.g., processor or controller 2 of Fig.1) may control a first pumping mechanism (e.g., inlet pumping mechanism 110 of Fig. 2) to draw blood material, associated with a specific subject, from a blood tube (e.g., tube 20 of Fig. 2) into a microfluidic chip (e.g., 150 of Fig. 2).

[0100] As shown in step S1010, the at least one processor a may acquire, via an imaging device (e.g., 160 of Fig. 2), at least one image (e.g., 160M of Fig. 2). Image 160M may depict blood cells in microfluidic chip 150.

[0101] As shown in step S1015, the at least one processor 2 may control a Raman spectroscopy device (e.g., 140 of Fig. 2) based on the at least one image 160M. Processor 2 may thus obtain one or more Raman spectra (e.g., 140SP of Fig. 2) from at least one cell of interest, pertaining to a predetermined cell type.

[0102] As shown in step S 1020, the at least one processor 2 may analyze the one or more Raman spectra 140SP, to determine a subtype (e.g., 170ST of Fig. 2) of the at least one cell of interest.

[0103] Reference is now made to Fig. 4, which is a flow diagram, depicting an example of analysis of blood material by system 10 according to some embodiments of the invention. As shown in Fig. 4, at least one processor or controller (e.g., controller 2 of Fig. 1) may employ system 10 to perform parallel analysis of blood samples, in a plurality (in this example - three) branches, using a respective plurality of microfluidic chips 150.

[0104] In the example of Fig. 4, the blood material analysis process begins at step S2010, where a queue manager is implemented. This step may be performed by processor 2, to manage the allocation of resources and selects blood tubes 20 for analysis.

[0105] In step S2020, system 10 may enter a tube ready state, where processor 2 may employ one or more actuators retrieve at least one blood tube 20 (e.g., from a stack of tubes) preparing that tube 20 for examination.

[0106] In step S2030, controller 2 may employ inlet pumping mechanism 110 to perform blood aspiration, drawing blood material 25 from the blood tube 20.

[0107] The analysis process may then split into a plurality (e.g., three) parallel branches, each corresponding to a different hardware, e.g., separate microfluidic chips 150. Each branch follows a similar sequence of steps. The following description will elaborate on steps of the first (leftmost) branch as an example.

[0108] Step S2100 may include a process of blood dilution. Controller 2 may perform this step by controlling the inlet pumping mechanism 110 to mix the drawn blood material 25 with a diluent such as PBS (Phosphate Buffered Saline), which may be a water-based salt solution that helps maintain a constant pH and osmotic balance. The dilution process may prepare the blood sample for analysis by adjusting its concentration and properties, as known in the art.

[0109] In step S2110, the diluted blood may be used to fill the microfluidic chip 150. controller 2 may carry out this operation by controlling inlet pumping mechanism 110, to pump the diluted blood into a microfluidic chip 150 of the first branch.

[0110] In step S2120, controller 2 may ensure sample stabilization, allowing the blood material 25 to settle within the microfluidic chip 150, preparing it for analysis. The specific settling mechanism may depend on the design of the microfluidic chip 150 and the requirements of the subsequent analysis steps.

[0111] The settling of blood material 25 within the microfluidic chip 150 may be achieved through various mechanisms. For example, the controller 2 may implement gravity-based sedimentation by pausing the flow within the microfluidic chip 150 for a predetermined time period, allowing the blood cells to naturally settle due to gravitational forces. Alternatively, the controller 2 may adjust the inlet pumping mechanism 110 to reduce the flow rate to a very low level, enabling cells to gradually settle in specific regions of the microfluidic chip 150. Acoustic focusing may also be employed, where the system incorporates acoustic transducers that generate standing waves within the microfluidic chip 150, gently pushing cells into predetermined positions or planes. In some cases, the dilution step may include chemical additives that modify the fluid properties to enhance cell settling without affecting subsequent analysis. The specific settling mechanism used may depend on the design of the microfluidic chip 150 and the requirements of the subsequent analysis steps.

[0112] Controller 2 may monitor the settling process, e.g., by using feedback from the imaging device 160 to ensure optimal sample preparation for Raman spectroscopy and other analytical techniques.

[0113] In step S2130, the controller may perform the acquisition flow, which is further elaborated herein, e.g., in relation to Fig. 5. During this step, imaging device 160 may capture cell images 160M, Raman spectroscopy device 140 may obtain Raman spectra 140SP from cells in the microfluidic chip 150. System 10 may then analyze images 160M, Raman spectra 140SP and / or CBC report 30CR to classify cell subtypes 170ST and / or subject conditions 190SC as elaborated herein (e.g., in relation to Fig. 2).

[0114] In step S2140, a cleaning flow is initiated. Controller 2 may conduct this operation by controlling the outlet pumping mechanism 120 (and / or inlet pumping mechanism 110), to flush the microfluidic chip 150, e.g., with a cleaning solution. This process is further elaborated herein in relation to Fig. 6.

[0115] Step S2150 indicates that the system is ready for a new sample. At this point, controller 2 may reiterate the process to prepare the microfluidic chip 150 for receiving another blood sample.

[0116] Other (e.g., second and third) branches of system 10 may followthe same sequence of operations (steps S2200-S2250 and S2300-S2350) as the first branch, but are performed on separate microfluidic chips 150. This parallel processing may allow system 10 to analyze multiple blood samples simultaneously, to increasing analysis throughput.

[0117] After each branch completes its cycle, the process may return to the queue manager at step S2010. The queue manager may then allocate the next blood tube 20 for analysis, continuing the cycle. This cyclical process may allow system 10 to continuously analyze blood samples, leveraging its parallel processing capabilities to efficiently characterize cells and determine their subtypes using a combination of spectroscopic and imaging techniques, as elaborated herein.

[0118] Reference is now made to Fig. 5, which is a flow diagram, depicting an example of an acquisition process, to analyze blood material 25 by embodiments of system 10.

[0119] Step S3005: Go to Starting Position - Controller 2 may initiate the acquisition process by moving the system components to their starting positions. This may involve using actuators 130 to position the Raman spectroscopy device 140 and imaging device 160 at their initial locations relative to the microfluidic chip 150.

[0120] Step S3010: Focus Check - Controller 2 may instruct imaging device 160 to perform a focus check to ensure that imaging device 160 is properly focused on the sample within microfluidic chip 150. This may involve capturing one or more initial microscopy images 160M of the sample within the microfluidic chip 150. Controller 2 may analyze these images 160M using image processing algorithms to assess focus quality. The focus check may evaluate parameters such as image sharpness, contrast, and edge definition of cellular structures. In some cases, controller 2 may capture a series of images 160M at different focal planes, creating a z-stack, to determine the optimal focal position. The system may also utilize autofocus algorithms that analyze frequency content or gradient information in the images 160M to quantitatively determine the best focus. If the focus check indicates suboptimal image quality, the process may proceed to a focus adjustment step.

[0121] Step S3015: Focus Adjustment - If the focus check fails, controller 2 may direct imaging device 160 to adjust its focus. This may include, for example, using actuators 130 to finetune the position of the imaging device 160 relative to the microfluidic chip 150, and / or adjusting inherent properties or positions of imaging device 160 for optimal focus.

[0122] Step S3020: Image Field of View (FOV) - Once focus is confirmed, controller 2 commands the imaging device 160 to capture an image 160M of the current field of view within the microfluidic chip 150.

[0123] Step S3025: Detect Relevant Cells - Controller 2 may utilize image analyzer 180 and its object recognizer 182 to process the captured image 160M. The Al-based object recognition algorithm 182 may identify relevant cells and their types 182T within the field of view.

[0124] Step S3030: Calculate Optimal Movement Pattern - Based on the detected cells, controller 2 may calculates an optimal movement pattern (also referred to as transition pattern) 130T for the Raman spectroscopy device 140 to efficiently analyze the identified cells of interest.

[0125] Step S3035: Move to Cell of Interest - Controller 2 uses actuators 130 to move the Raman spectroscopy device 140 according to the calculated transition pattern 130T, positioning it at the location of a cell of interest within the microfluidic chip 150.

[0126] Step S3040: Raman Focus Check - Controller 2 may instruct the Raman spectroscopy device 140 to perform a focus check, ensuring it is properly focused on the cell of interest.

[0127] Step S3045: Raman Focus Adjustment - If the Raman focus check fails, controller 2 may direct actuators 130 to adjust the position of the Raman spectroscopy device 140 to achieve proper focus on the cell of interest. Additionally, or alternatively, controller 2 may direct inherent components of Raman spectroscopy device 140 to adjust so as to obtain optimal focus.

[0128] Step S3050: Acquire Raman Spectra - Once focus is confirmed, controller 2 may command Raman spectroscopy device 140 to acquire Raman spectra 140SP from the cell of interest. As explained herein, subtype classifier 170 may then analyze these spectra, potentially in combination with the cell image 160M, to classify the cell subtype 170ST.

[0129] Step S3055: Check if the Last Cell in FOV - Controller 2 may determine whether the current cell is the last cell of interest in the field of view. If not, the process may return to step S3035 to move to the next cell of interest.

[0130] Step S3060: Check if Desired Cell Number Reached - If the last cell in the field of view has been analyzed, controller 2 may check if the desired number of cells of interest has been reached for the current sample.

[0131] Step S3065: Finish Sample and Clean Chip - If the desired cell number is reached, controller 2 may initiate a process to finish the sample analysis. This may involve using the outlet pumping mechanism 120 to flush the microfluidic chip 150 with a cleaning solution, as elaborated herein in relation to Fig. 2 and Fig. 6.

[0132] Step S3070: Check for Additional FOVs - If the desired cell number has not been reached, controller 2 may check whether there are additional fields of view available in the current microfluidic chip 150.

[0133] Step S3075: Move XY Stage to New FOV - If such additional fields of view are available, controller 2 may use actuators 130 to move an XY stage, to reposition microfluidic chip 150 relative to the imaging device 160 and / or Raman spectroscopy device 140, to analyze a new field of view.

[0134] Step S3080: Fill Chip - If no more fields of view are available in the current chip, controller 2 may instruct the inlet pumping mechanism 110 to fill the microfluidic chip 150 with a new sample of blood material 25 from blood tube 20.

[0135] Step S3085: Sample Stabilization - After filling the chip, controller 2 may allow time for sample stabilization, which may involve processes described earlier for settling the blood material 25 within the microfluidic chip 150.

[0136] The process may then return to the starting position (step S3005) to begin analysis of the new sample or field of view. Throughout this process, controller 2 may interface with the CBC module 30 to integrate CBC report data 3 OCR with the Raman 140SP and imaging analysis results 160M and may utilize the subject classifier 190 to determine the subject's condition 190SC based on the accumulated data.

[0137] Fig. 6 illustrates a flowchart depicting an example of a cleaning process for a microfluidic subsystem of system 10. This process may be managed by at least one processor 2, according to some embodiments of the invention.

[0138] The cleaning process begins at step S4005, where controller 2 may initiate the cleaning sequence by returning system components to their initial state or position. This may involve using actuators 130 to reposition the Raman spectroscopy device 140 and imaging device 160 away from the microfluidic chip 150, ensuring unobstructed fluid flow through the chip during cleaning.

[0139] In step S4010, controller 2 may activate inlet pumping mechanism 110 to flow a first cleaning fluid through the microfluidic chip 150 at a flow speed of Fl for a duration of Ti l. This initial cleaning fluid may be designed to remove cellular debris and other residues from the previous blood sample analysis.

[0140] Following the first cleaning fluid flow, step S4015 may involve a waiting period of T12. During this time, controller 2 may maintain the microfluidic chip 150 in a static state, allowing the cleaning fluid to soak and further dissolve any remaining contaminants.

[0141] The process then continues to step S4020, where controller 2 may activate inlet pumping mechanism 110 again to flow a second cleaning fluid through the system at a flow speed of F2 for a duration of T21. This second fluid may have different chemical properties from the first, targeting specific types of residues or contaminants.

[0142] After the second cleaning fluid, step S4025 may involve another waiting period of T22, managed by controller 2. This pause may allow for any chemical reactions between the cleaning fluid and contaminants to complete.

[0143] In step S4030, controller 2 may initiate the flow of a third cleaning fluid through the system at a flow speed of F3 for a duration of T31. Following the third cleaning fluid, step S4035 may involve a waiting period of T32, again controlled by controller 2. This final soak may ensure thorough cleaning of the microfluidic chip 150.

[0144] The process then proceeds to step S4040, where controller 2 may activate inlet pumping mechanism 110 to flow PBS (Phosphate Buffered Saline) through the system at a flow speed of F4 for a duration of T41. This PBS rinse may serve to remove any remaining cleaning fluids and prepare the microfluidic chip 150 for the next blood sample analysis.

[0145] Step S4045 may involve a waiting period of T42, completing the cleaning sequence. During this time, controller 2 may perform system checks, such as verifying the cleanliness of the microfluidic chip 150 using the imaging device 160.

[0146] Throughout this process, controller 2 may monitor and adjust the flow rates and durations as needed, potentially using feedback from the imaging device 160 or other sensors to ensure thorough cleaning. The outlet pumping mechanism 120 may be used to remove waste fluids, while the actuators 130 may be employed to reposition components as necessary during the cleaning process.

[0147] This comprehensive cleaning protocol, managed by controller 2, ensures that the microfluidic subsystem of system 10 is thoroughly cleaned and prepared for subsequent blood sample analyses, maintaining the accuracy and reliability of the system's analytical capabilities.

[0148] Embodiments of the invention may provide significant advancements in blood analysis by integrating multiple analytical techniques into a single, automated system 10. This integrationmay allow for more comprehensive, efficient, and accurate blood sample characterization, potentially improving diagnostic capabilities and clinical decision-making.

[0149] Additionally, system 10 capability of selecting cells of interest, and applying a combination of analytical processes (e.g., Al based image analysis and Raman spectra analysis) on the same cells of interest, allow system 10 to identify various cell types and subtypes, thereby facilitating nuanced understanding of a patient's immune status and overall health.

[0150] The selection of cells of interest may be based on specific use cases for which system 10 may be employed. For example, specific types of white blood cells (WBCs) may be selected as cells of interest when they are known to be indicative of a suspected condition of the subject. As explained herein, image analysis module 180 (object recognizer 182) may identify these cells of interest (cell types 182T) based, for example, on morphological features, size, or other distinguishing characteristics visible in the captured images 160M. Subsequently, system 10 may analyze these selected cells to identify cell subtypes 170ST using the subtype classifier 170, which may integrate data from both imaging and Raman spectroscopy. In some cases, the subject classifier 190 may utilize the cell subtype information, along with other data such as CBC report 30CR and / or Laboratory Information System (LIS) information 40, to determine the subject's condition 190SC. This targeted approach may allow system 10 to focus its analytical capabilities on the most relevant cellular components for each specific use case, potentially enhancing the efficiency and accuracy of the analysis.

[0151] For example, system 10 may be used to analyze blood material 25 for the following purposes:

[0152] Comprehensive WBC Analysis: System 10 may perform a detailed analysis of various white blood cell types, including lymphocytes, granulocytes, and monocytes. The imaging device 160 may capture high-resolution images 160M of cells in the microfluidic chip 150. The image analyzer 180 and object recognizer 182 may then identify and classify different cell types (e.g., lymphocytes, granulocytes, and monocytes) based on their morphological features 185. The Raman spectroscopy device 140 may obtain Raman spectra 140SP from these identified cells of interest, providing molecular-level information.

[0153] Subtype classifier 170 may integrate this data to determine specific cell subtypes 170ST and their relative percentage or proportions 170PC. This comprehensive analysis may aidin detecting abnormalities across multiple WBC populations, potentially indicating various health conditions or diseases.

[0154] Immune System Profiling: System 10 may be useful for detailed immune system profding. For example, for lymphocytes, the subtype classifier 170 may determine the proportions of T-cells, B-cells, and NK cells (TBNK analysis). Additionally, it may analyze the activation states of these cells based on their Raman spectra 140SP. For granulocytes and monocytes, the system may assess their activation levels and functional states. This comprehensive immune profde may be valuable for monitoring immune-related disorders, assessing response to immunotherapies, or evaluating overall immune system health.

[0155] Hematological Malignancy Screening: In screening for hematological malignancies, system 10 may analyze the morphological features 185 and molecular characteristics of blood cells. The imaging device 160 and image analyzer 180 may detect abnormal cell morphologies, while the Raman spectroscopy device 140 may identify molecular changes associated with malignant transformations. The subtype classifier 170 may integrate this information to flag potentially malignant cells or abnormal cell populations. This capability may aid in early detection or monitoring of conditions such as leukemias or lymphomas.

[0156] System 10 may be effective in detecting and characterizing various hemoglobinopathies and related conditions. For instance, the Raman spectroscopy device 140 may obtain unique spectral signatures from red blood cells that can distinguish between normal adult hemoglobin, fetal hemoglobin, and abnormal hemoglobin variants such as those found in sickle cell disease. The image analyzer 180 and object recognizer 182 may identify morphological changes associated with these conditions, such as the characteristic sickle shape in sickle cell disease or the microcytic, hypochromic appearance of red blood cells in iron deficiency anemia. The subtype classifier 170 may integrate this spectroscopic and morphological data to categorize different types of hemoglobinopathies. By combining these results with CBC data from module 30, the subject classifier 190 may provide a comprehensive assessment of the patient's hemoglobin status, potentially aiding in the diagnosis and monitoring of conditions such as sickle cell disease, thalassemia, and iron deficiency anemia. This capability may enable more precise and early detection of these disorders, potentially improving patient management and treatment outcomes.

[0157] Sepsis Detection: For early sepsis detection, system 10 may analyze changes in WBC populations and their activation states. The object recognizer 182 may identify an increase inneutrophil count, while the Raman spectroscopy device 140 and subtype classifier 170 may detect changes in neutrophil activation states. Simultaneously, the system may assess lymphocyte counts and subtypes. By integrating this information with data from the CBC module 30, the subject classifier 190 may provide an early indication of sepsis risk.

[0158] System 10 may also be utilized for assessing Systemic Inflammatory Response Syndrome (SIRS) and monitoring progression towards septic shock. The comprehensive analysis of white blood cell populations, their activation states, and molecular characteristics may provide valuable insights into the systemic inflammatory response.

[0159] In cases of SIRS, system 10 may detect alterations in white blood cell counts, particularly an increase in neutrophils or a decrease in lymphocytes. The Raman spectroscopy device 140 may identify changes in cellular metabolism and activation states associated with the inflammatory response. The subtype classifier 170 may analyze these spectral signatures to determine the degree of cellular activation and stress.

[0160] For patients at risk of progressing from sepsis to septic shock, system 10 may provide continuous monitoring of immune cell function and activation. The image analyzer 180 and object recognizer 182 may track changes in cell morphology that indicate increased cellular stress or dysfunction. The Raman spectroscopy device 140 may detect molecular changes associated with cellular energy depletion and oxidative stress, which are hallmarks of progression towards septic shock.

[0161] By integrating data from the CBC module 30 with the detailed cellular analysis, the subject classifier 190 may generate a risk score for septic shock progression. This score may take into account factors such as the ratio of immature to total neutrophils, changes in lymphocyte subpopulations, and alterations in cellular activation states as determined by Raman spectroscopy.

[0162] Anemia Characterization: In characterizing anemia, system 10 may analyze red blood cell (RBC) properties alongside WBC analysis. The imaging device 160 and image analyzer 180 may assess RBC morphology, while the Raman spectroscopy device 140 may provide information about hemoglobin content and structure. This data, combined with CBC results from module 30, may allow the subject classifier 190 to differentiate between various types of anemia, such as iron deficiency anemia or hemoglobinopathies.

[0163] System 10 may be capable of detecting and analyzing the presence of red blood cell (RBC) fragments in blood material 25, which can be indicative of various hematologicalconditions such as hemolytic anemia or microangiopathic hemolytic disorders. The imaging device 160 may capture high-resolution images 160M of cells and cell fragments within the microfluidic chip 150. The image analyzer 180, utilizing advanced image processing algorithms, may identify and characterize RBC fragments based on their unique morphological features 185, such as irregular shapes, smaller sizes compared to intact RBCs, or the presence of schistocytes (fragmented RBCs with sharp angles). The object recognizer 182 may be trained to distinguish these fragments from other cellular components or debris. Additionally, the Raman spectroscopy device 140 may obtain Raman spectra 140SP from these fragments, potentially providing molecular information about their composition and origin. The subtype classifier 170 may integrate the imaging and spectroscopic data to categorize different types of RBC fragments and quantify their prevalence. By correlating this information with data from the CBC module 30, the subject classifier 190 may assist in diagnosing specific conditions associated with RBC fragmentation, such as thrombotic thrombocytopenic purpura (TTP), disseminated intravascular coagulation (DIC), or mechanical hemolysis due to prosthetic heart valves. The ability of system 10 to detect and characterize RBC fragments may enhance the diagnostic capabilities for a range of hematological disorders and provide valuable insights into the underlying mechanisms of RBC destruction.

[0164] System 10 may provide valuable insights into distinguishing between viral and bacterial infections by analyzing the immune system activation profde, particularly focusing on lymphocytes and neutrophils. The imaging device 160 and object recognizer 182 may identify and quantify different types of white blood cells, while the Raman spectroscopy device 140 may obtain detailed molecular information about their activation states. For viral infections, the subtype classifier 170 may detect an increase in activated lymphocytes, particularly T-cells, and their specific subtypes (e.g., CD8+ cytotoxic T-cells). In contrast, bacterial infections may be characterized by a marked increase in activated neutrophils, which the system can identify through both morphological changes (detected by the image analyzer 180) and specific molecular signatures in their Raman spectra 140SP. The subject classifier 190 may integrate this information with data from the CBC module 30, such as the neutrophil-to-lymphocyte ratio, to provide a comprehensive assessment of the immune system's response. This analysis may aid in differentiating between viral and bacterial infections.

[0165] System 10 may also be utilized for detecting the presence of blast cells in blood material 25, which can be indicative of various hematological disorders, including acute leukemias. The imaging device 160 may capture high-resolution images 160M of cells in the microfluidic chip 150, while the image analyzer 180 and object recognizer 182 may identify cells with morphological features 185 characteristic of blast cells, such as large size, high nucleus-to- cytoplasm ratio, and the presence of nucleoli. Simultaneously, the Raman spectroscopy device 140 may obtain Raman spectra 140SP from these cells, providing molecular-level information that may distinguish blast cells from mature blood cells. The subtype classifier 170 may integrate the imaging and spectroscopic data to accurately identify and quantify blast cells. By combining this analysis with CBC data from module 30, the subject classifier 190 may assess the likelihood of conditions such as acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL). This capability of system 10 to detect and characterize blast cells may aid in early diagnosis, disease monitoring, and evaluation of treatment response in hematological malignancies.

[0166] In all these examples, system 10 may benefit from the integration of multiple analytical techniques, including Al-based image analysis and Raman spectroscopy applied to the same physical cells of interest, optionally in conjunction with additional information such as traditional CBC analysis.

[0167] This unique combination may allow system 10 to provide a comprehensive and nuanced analysis of blood material 25. This may lead to improved diagnostic accuracy and more personalized patient care.

[0168] As known in the art, CBC module 30 may provide standard blood count data that is typically used in clinical settings. However, beyond the standard CBC parameters, there are additional measurements and analyses that can be performed on blood samples. These "research parameters" may provide more detailed or specialized information about the blood cells and their characteristics. Such research parameters may not have yet been approved by the Food and Drug Administration (FDA) for routine clinical use, and may be considered experimental or investigational, but may nevertheless include additional information that may be valuable for determining a fine-grained classification of a condition of a subject (e.g., a medical condition of a patient).

[0169] By integrating such research parameters with the standard CBC data and the advanced analysis capabilities of system 10 elaborated herein (including combining Raman spectroscopyanalysis and Al-driven image analysis on the same cells), system 10 may provide a more comprehensive and detailed analysis of blood samples. Such integration may allow researchers and clinicians to explore correlations between standard CBC data, advanced spectroscopic measurements, and these additional research parameters, providing new diagnostic or prognostic indicators.

[0170] Additionally, or alternatively, System 10 may be designed to interface with data from a Laboratory Information System (LIS) 40. As known in the art, LIS information 40 typically contains a wide range of patient information beyond just blood test results, which may include, for example, medical history, other laboratory tests, demographic data, clinical notes, and the like.

[0171] According to some embodiments, system 10 may introduce LIS information as additional input to subject classification module 190, e.g., in addition to subtype classification 170ST and / or CBC report data 30CR. Subject classification module 190 may thereby incorporate additional patient parameters into its analysis, potentially improving the accuracy and context of its prediction of a subject's condition 190SC.

[0172] Additionally, or alternatively, subject classifier 190 may use this additional information from the LIS 40 to identify correlations between blood analysis results (e.g., 3 OCR, 170ST) and other patient parameters, potentially uncovering new diagnostic indicators or patterns.

[0173] Additionally, system 10 may utilize LIS information 40 to enable personalized analysis and interpretation of blood test results, taking into account individual patient factors and history. Embodiments of the invention may thus provide a comprehensive and contextualized analysis of blood samples, potentially leading to more accurate diagnoses and personalized treatment strategies.

[0174] Unless explicitly stated, the method embodiments described herein are not constrained to a particular order or sequence. Furthermore, all formulas described herein are intended as examples only and other or different formulas may be used. Additionally, some of the described method embodiments or elements thereof may occur or be performed at the same point in time.

[0175] While certain features of the invention have been illustrated and described herein, many modifications, substitutions, changes, and equivalents may occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.

[0176] Various embodiments have been presented. Each of these embodiments may of course include features from other embodiments presented, and embodiments not specifically described may include various features described herein.

Claims

CLAIMS1. A system for analyzing blood material, the system comprising: a first pumping mechanism; a microfluidic chip; an imaging device; a Raman spectroscopy device; a non-transitory memory device, wherein modules of instruction code are stored; and at least one processor, wherein the at least one processor is configured to execute the modules of instruction code to: control the first pumping mechanism to draw blood material, associated with a specific subject, from a blood tube into the microfluidic chip; acquire, via the imaging device, at least one image, depicting cells in the microfluidic chip; based on the at least one image, control the Raman spectroscopy device to obtain one or more Raman spectra from a cell of interest, pertaining to a predetermined cell type; and analyze the one or more Raman spectra, to determine a subtype of the at least one cell of interest.

2. The system of claim 1, wherein the at least one processor is configured to: apply an object recognition algorithm on the at least one image, to identify a plurality of cells of interest, as pertaining to the predetermined cell type; based on said identification, calculate a transition pattern of the Raman spectroscopy device; control one or more actuators, to adjust a position of the Raman spectroscopy device based on the transition pattern; and control the Raman spectroscopy device to specifically illuminate one or more cells of the plurality of cells of interest, so as to obtain the one or more Raman spectra from each of the one or more cells of interest.

3. The system according to any one of claims 1 -2, wherein the at least one processor is configured toreceive a dataset comprising a plurality of Raman spectra associated with a respective plurality of cells; receive a plurality of subtype annotations, respectively associated with the plurality of cells, wherein each subtype annotation indicates a specific cell subtype; and use the one or more subtype annotations as supervisory information, to train a Machine Learning (ML)-based, subtype classification model, to predict a subtype of an incident cell, based on the received dataset.

4. The system of claim 3, wherein the at least one processor is configured to: receive one or more Raman spectra of a target cell; provide the one or more Raman spectra as input to the subtype classification model; and obtain, from the subtype classification model, a subtype of the target cell, based on said training.

5. The system according to any one of claims 1 -4, wherein the at least one processor is configured to receive a dataset comprising: (a) a plurality of Raman spectra associated with a respective plurality of cells, and (b) a plurality of images, respectively depicting the plurality of cells; receive a plurality of subtype annotations, respectively associated with the plurality of cells, wherein each annotation indicates a specific cell subtype; and use the one or more subtype annotations as supervisory information, to train an ML-based, subtype classification model to predict a subtype of an incident cell, based on the received dataset.

6. The system of claim 5, wherein the at least one processor is configured to: receive one or more Raman spectra of a target cell; provide the one or more Raman spectra as input to the subtype classification model; receive, via the imaging device, an image depicting the target cell; provide said image of the target cell as input to the subtype classification model; and obtain, from the subtype classification model, a subtype of the target cell of interest, based on said training.

7. The system according to any one of claims 5-6, wherein the at least one processor is further configured to for each image of the plurality of images, apply an image analysis algorithm, to obtain at least one morphological feature of the respective depicted cell; use the one or more subtype annotations as supervisory information, to train an ML-based, subtype classification model to predict a subtype of an incident cell, based on (a) the plurality of Raman spectra of the dataset, and (b) the at least one morphological feature of the plurality of images.

8. The system of claim 7, wherein the at least one processor is configured to: receive one or more Raman spectra of a target cell; provide the one or more Raman spectra as input to the subtype classification model; receive, via the imaging device, an image depicting the target cell; apply the image analysis algorithm on said image, to obtain at least one morphological feature of the depicted target cell; provide the at least one morphological feature as input to the subtype classification model; and obtain, from the subtype classification model, a subtype of the target cell, based on said training.

9. The system according to any one of claims 1 -8, wherein the at least one processor is configured to: determining a subtype of a plurality of cells in the blood material of the subject; for each cell subtype, compute a subtype percentage, representing percentage of that subtype among cells of the same type; for each cell subtype, evaluate a subtype quantity value, representing a number of cells of that subtype in the at least one image; and determine a condition of the subject based on (i) said subtype percentage and (ii) said subtype quantity value.

10. The system of claim 9, associated with a Complete Blood Count (CBC) module, wherein the at least one processor is further configured to: receive, from the CBC module, at least one CBC report associated with the subject; provide the at least one CBC report as a first input to a pretrained, ML-based, subject classification model; provide at least one of (i) a subtype percentage of at least one subtype and (ii) a subtype quantity value of at least one subtype, as a second input to the subject classification model; and obtain, from the subject classification model, a prediction of the condition of the subject, based on the first and second input.

11. The system of claim 10, wherein the at least one processor is configured to: receive a second dataset comprising (i) a plurality subtype percentages, associated with a cohort of subjects, (ii) a plurality of subtype quantity values associated with the cohort of subjects, and (iii) a plurality of CBC report data elements associated with the cohort of subjects; receive a plurality of condition annotations, respectively associated with each of the cohort of subjects, wherein each condition annotation indicates a medical condition of the respective subject; and use the plurality of condition annotations, to train the ML-based subject classification model to predict the medical condition of an incident subject, based on the second dataset.

12. The system according to any one of claims 1-11, further comprising a second pumping mechanism, and wherein the at least one processor is configured to: obtain a predetermined number of Raman spectra, corresponding to a predetermined number of cells in the blood material; and controlling the second pumping mechanism to flush a cleaning solution via the microfluidic chip, thereby preparing the microfluidic chip to receive blood material from another blood tube.

13. The system according to any one of claims 1-12, wherein the first pumping mechanism comprises a puncturing device, a first pump and an inlet valve, and wherein the at least one processor is configured to: control the puncturing device, to enter the blood tube;control the first pump, to draw blood material from the blood tube via the puncturing device; open the inlet valve, to insert a sample of the blood material into the microfluidic chip; and close the inlet valve, to stabilize the sample of blood material in the microfluidic chip.

14. The system according to any one of claims 5-13, wherein the at least one processor is configured to control the imaging device, so as to: auto-focus on the stabilized sample of blood, in a segment of the microfluidic chip; and acquire at least one first image depicting cells in the segment of the microfluidic chip.

15. The system according to any one of claims 1-14, further comprising an automated CBC module, wherein the at least one processor is configured to: obtain, from the CBC module, at least one CBC report selected from (i) a hematology analysis report and (ii) a blood-chemistry analysis report; identify at least one anomaly in the subject’s blood material based on the CBC report data element; and based on the identification of said at least one anomaly, control at least one of the Raman spectroscopy device and first pumping device, to obtain the Raman spectrum.

16. A method of analyzing blood material by at least one processor, the method comprising: controlling a first pumping mechanism to draw blood material, associated with a specific subject, from a blood tube into a microfluidic chip; acquiring, via an imaging device, at least one image depicting cells in the microfluidic chip; based on the at least one image, controlling a Raman spectroscopy device to obtain one or more Raman spectra from a cell of interest, pertaining to a predetermined cell type; and analyzing the one or more Raman spectra, to determine a subtype of the at least one cell of interest.

17. The method of claim 16, further comprising: applying an object recognition algorithm on the at least one image, to identify a plurality of cells of interest, as pertaining to the predetermined cell type;based on said identification, calculating a transition pattern of the Raman spectroscopy device; controlling one or more actuators, to adjust a position of the Raman spectroscopy device based on the transition pattern; and controlling the Raman spectroscopy device to specifically illuminate one or more cells of the plurality of cells of interest, so as to obtain the one or more Raman spectra from each of the one or more cells of interest.

18. The method according to any one of claims 16-17, further comprising: receiving a dataset comprising a plurality of Raman spectra associated with a respective plurality of cells; receiving a plurality of subtype annotations, respectively associated with the plurality of cells, wherein each subtype annotation indicates a specific cell subtype; and using the one or more subtype annotations as supervisory information, to train a Machine Learning (ML)-based, subtype classification model, to predict a subtype of an incident cell, based on the received dataset.

19. The method of claim 18, further comprising: receiving one or more Raman spectra of a target cell; providing the one or more Raman spectra as input to the subtype classification model; and obtaining, from the subtype classification model, a subtype of the target cell, based on said training.

20. The method according to any one of claims 16-19, further comprising: receiving a dataset comprising: (a) a plurality of Raman spectra associated with a respective plurality of cells, and (b) a plurality of images, respectively depicting the plurality of cells; receiving a plurality of subtype annotations, respectively associated with the plurality of cells, wherein each annotation indicates a specific cell subtype; andusing the one or more subtype annotations as supervisory information, to train an ML- based, subtype classification model to predict a subtype of an incident cell, based on the received dataset.

21. The method of claim 20, further comprising: receiving one or more Raman spectra of a target cell; providing the one or more Raman spectra as input to the subtype classification model; receiving, via the imaging device, an image depicting the target cell; providing said image of the target cell as input to the subtype classification model; and obtaining, from the subtype classification model, a subtype of the target cell of interest, based on said training.

22. The method according to any one of claims 20-21, further comprising: for each image of the plurality of images, applying an image analysis algorithm, to obtain at least one morphological feature of the respective depicted cell; using the one or more subtype annotations as supervisory information, to train an ML- based, subtype classification model to predict a subtype of an incident cell, based on (a) the plurality of Raman spectra of the dataset, and (b) the at least one morphological feature of the plurality of images.

23. The method of claim 22, further comprising: receiving one or more Raman spectra of a target cell; providing the one or more Raman spectra as input to the subtype classification model; receiving, via the imaging device, an image depicting the target cell; applying the image analysis algorithm on said image, to obtain at least one morphological feature of the depicted target cell; providing the at least one morphological feature as input to the subtype classification model; and obtaining, from the subtype classification model, a subtype of the target cell, based on said training.

24. The method according to any one of claims 16-23, further comprising: determining a subtype of a plurality of cells in the blood material of the subject; for each cell subtype, computing a subtype percentage, representing percentage of that subtype among cells of the same type; for each cell subtype, evaluating a subtype quantity value, representing a number of cells of that subtype in the at least one image; and determining a condition of the subject based on (i) said subtype percentage and (ii) said subtype quantity value.

25. The method of claim 24, further comprising: receiving, from a Complete Blood Count (CBC) module, at least one CBC report associated with the subject; providing the at least one CBC report as a first input to a pretrained, ML-based, subject classification model; providing at least one of (i) a subtype percentage of at least one subtype and (ii) a subtype quantity value of at least one subtype, as a second input to the subject classification model; and obtaining, from the subject classification model, a prediction of the condition of the subject, based on the first and second input.

26. The method of claim 25, further comprising: receiving a second dataset comprising (i) a plurality subtype percentages, associated with a cohort of subjects, (ii) a plurality of subtype quantity values associated with the cohort of subjects, and (iii) a plurality of CBC report data elements associated with the cohort of subjects; receiving a plurality of condition annotations, respectively associated with each of the cohort of subjects, wherein each condition annotation indicates a medical condition of the respective subject; and using the plurality of condition annotations, to train the ML-based subject classification model to predict the medical condition of an incident subject, based on the second dataset.

27. The method according to any one of claims 16-26, further comprising:obtaining a predetermined number of Raman spectra, corresponding to a predetermined number of cells in the blood material; and controlling a second pumping mechanism to flush a cleaning solution via the microfluidic chip, thereby preparing the microfluidic chip to receive blood material from another blood tube.

28. The method according to any one of claims 16-27, wherein the first pumping mechanism comprises a puncturing device, a first pump and an inlet valve, and wherein the method further comprises: controlling the puncturing device, to enter the blood tube; controlling the first pump, to draw blood material from the blood tube via the puncturing device; opening the inlet valve, to insert a sample of the blood material into the microfluidic chip; and closing the inlet valve, to stabilize the sample of blood material in the microfluidic chip.

29. The method according to any one of claims 20-28, further comprising controlling the imaging device, so as to: auto-focus on the stabilized sample of blood, in a segment of the microfluidic chip; and acquire at least one first image depicting cells in the segment of the microfluidic chip.

30. The method according to any one of claims 16-29, further comprising: obtaining, from an automated CBC module, at least one CBC report selected from (i) a hematology analysis report and (ii) a blood-chemistry analysis report; identifying at least one anomaly in the subject's blood material based on the CBC report data element; and based on the identification of said at least one anomaly, controlling at least one of the Raman spectroscopy device and first pumping device, to obtain the Raman spectrum.

Images