System and method for correcting patient indices - Patents.com

Abstract

A system and / or method for determining the immune activation state of a subject includes deforming a white blood cell in a microfluidic channel, acquiring a plurality of images of the white blood cell, determining a biophysical parameter of the white blood cell, adjusting the biophysical parameter, and determining the immune activation state of the subject based on the biophysical parameter.

Description

[Technical field]

[0001] REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Patent Application No. 17 / 592,040, filed February 3, 2022, and U.S. Patent Application No. 17 / 841,519, filed June 15, 2022, each of which is incorporated by reference in its entirety.

[0002] The present invention relates generally to the health field, and more specifically to novel and useful systems and methods in the health field. [Brief description of the drawings]

[0003] [Figure 1] FIG. 1 is a schematic diagram of the apparatus. [Diagram 2] FIG. 1 is a schematic diagram of the method. [Diagram 3] FIG. 2 is a schematic diagram of an example of a range of flow rates over which adjustments are applied. [Figure 4] FIG. 13 is a schematic diagram of an example of adjusting measurements based on flow rate. [Figure 5A] 5A-5C are schematic diagrams of examples of events that pass (5A) or fail (5B or 5C) through an event filter based on the position of the particle's inlet and / or outlet in a microfluidic channel. [Figure 5B] 5A-5C are schematic diagrams of examples of events that pass (5A) or fail (5B or 5C) through an event filter based on the position of the particle's inlet and / or outlet in a microfluidic channel. [Figure 5C] 5A-5C are schematic diagrams of examples of events that pass (5A) or fail (5B or 5C) through an event filter based on the position of the particle's inlet and / or outlet in a microfluidic channel. [Figure 6] FIG. 13 is a schematic diagram of an example of overlapping frames of particles with different handedness passing through a microfluidic channel, where measurements can be adjusted based on the handedness of the flow channel. [Figure 7A]FIG. 1 is a schematic diagram of an example feature detector for identifying overlapping particles. [Figure 7B] FIG. 1 is a schematic diagram of an example feature detector for identifying overlapping particles. [Figure 8] 1 is a schematic diagram of an exemplary z-plane in which a particle may reside, and adjustments to measurements may be z-plane dependent. [Figure 9] FIG. 1 is a schematic diagram of an example of determining a health status of a patient based on a dataset. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0004] The following description of preferred embodiments of the invention is not intended to limit the invention to those preferred embodiments, but rather to enable any person skilled in the art to make and use the invention.

[0005] overview 1, the system (10) may include an image acquisition system (200), a sample cartridge (100), a computing system (300), and / or any suitable components. The computing system (300) may include an image analyzer (320), a filter (340), a conditioner (360), an index calculator, a health scorer (380), and / or any suitable components or modules.

[0006] As shown in FIG. 2, the method (20) may include receiving a dataset associated with a patient (S100), filtering the dataset (S200), determining a metric from the dataset (S300), adjusting the metric (S400), determining an immune activation state of the patient (S500), providing an intervention to the patient (S600), and / or any suitable steps.

[0007] The systems and methods are preferably used to measure and detect biophysical properties of cells, but can measure properties of any particle and / or material (e.g., that fits into a microfluidic channel). The systems and / or methods can additionally or alternatively function to diagnose the health status (e.g., its immune activation state) of a subject (e.g., a patient). In a specific example, the systems and / or methods can be used in emergency departments (e.g., in hospitals), urgent care, and / or doctor's offices to triage patients (e.g., to diagnose and / or determine the probability that a patient is experiencing an illness, such as a sepsis-related illness). However, the systems and / or methods can be used in any manner.

[0008] profit. Variations of this technology can provide several benefits and / or advantages.

[0009] First, variations of the present technology can improve the reproducibility, repeatability, and / or reliability of patient sample analysis. The improved reproducibility, repeatability, and / or reliability can be across different practitioners, different medical procedures, different geographic regions, different times (e.g., hours, days, weeks, months, years, etc.), different systems, and / or under any suitable conditions. For example, the system and / or method can achieve a reproducibility of a score unit of 1.0 (e.g., the score can range from about 0 to 10, 0 to 100, and / or have any suitable range), and the score can be related to an immune activation state. In another example, the system and / or method can achieve a reproducibility of at least 10% (e.g., 0.01%, 0.05%, 0.1%, 0.5%, 1%, 5%, 10%, values ​​therebetween, etc.) in measuring a biophysical parameter of a sample in different conditions. Improved reproducibility, repeatability, and / or reliability can be achieved, for example, by adjusting biophysical parameters, images, scores, metrics, and / or any suitable data set or analysis thereof based on measured system and / or sample performance or other characteristics, however, the system and / or method can achieve any suitable reproducibility, repeatability, and / or reliability.

[0010] However, variations on the technology may impart any other suitable benefits and / or advantages.

[0011] As used herein, "substantially" or other similar terms (e.g., "about," "approximately," etc.) may be within a given error threshold or tolerance of a metric, component, or other reference (e.g., within 0.001%, 0.01%, 0.1%, 1%, 5%, 10%, 20%, 30%, etc. of the reference) or may be otherwise interpreted.

[0012] system. The system (10) is preferably operable to measure a data set associated with a patient, analyze the data set of the patient, determine the health status of the patient, adjust the data set and / or its analysis, and / or otherwise. The data set may include a set of images of the patient sample (15) (e.g., a plurality of images of each of a plurality of cells (17) of the patient sample, a plurality of images of each of a plurality of cells while the cells are subjected to deformation, stress, force, pressure, etc.), a set of biophysical properties (e.g., mechanical properties, trajectory parameters, etc.) of the patient sample, system properties (e.g., applied force, pressure, tension, flow rate, vorticity, size, etc.), and / or any suitable data. Patient samples may include white blood cells (e.g., a sample from a patient that may include a prepared sample disclosed in U.S. Patent Application No. 17 / 575,388, entitled "SYSTEM AND METHOD FOR DETERMINING AN IMMUNE ACTIVATION STATE," filed January 13, 2022, and / or U.S. Patent Application No. 17 / 575,422, entitled "SYSTEM AND METHOD FOR DETERMINING AN IMMUNE ACTIVATION STATE," filed January 13, 2022, each of which is incorporated herein by reference in its entirety, white blood cells from a patient, a sample consisting essentially of blood from a patient), red blood cells (erythrocytes), platelets (thrombocytes), plasma, reticulocytes, platelets, cellular debris, mechanical debris, dust, and / or any suitable sample associated with a patient (e.g., collected from a sample of blood, feces, urine, biopsy, etc.). For example, a sample may contain approximately 10,000 to 1,000,000 white blood cells (17) (and potentially red blood cells and / or red blood cell debris).

[0013] The sample (15) is preferably measured (e.g., tested, deformed, used, etc.) in a flow cytometer (e.g., a deformability cytometer, constriction-based deformability cytometry (cDC), shear flow deformability cytometry (sDC), and / or extensional flow deformability cytometry (xDC), etc.). However, the sample can additionally or alternatively be measured using an atomic force microscope, using an optical probe, and / or in any manner. For example, the sample can be loaded into a cartridge (100) (e.g., a microfluidic cartridge, a microfluidic cartridge, etc.) that is measured using flow cytometry.The cartridge may include any suitable microfluidic channel (150) or a plurality of microfluidic devices, as described in U.S. Application No. 16 / 374,663, filed April 3, 2019 and entitled "SYSTEM AND METHOD FOR DEFORMING AND ANALYZING PARTICLES," U.S. Application No. 15 / 868,025, filed January 11, 2018 and entitled "METHOD AND DEVICE FOR HIGH-THROUGHPUT CELL DEFORMABILITY MEASUREMENTS," U.S. Application No. 16 / 676,352, filed November 6, 2019 and entitled "METHOD AND DEVICE FOR HIGH-THROUGHPUT SOLUTION EXCHANGE FOR CELL AND PARTICLE SUSPENSIONS," U.S. Application No. 16 / 676,352, filed October 18, 2013 and entitled "SYSTEM AND METHOD FOR No. 9,464,977, filed March 3, 2017, and entitled "SYSTEM AND METHOD FOR DEFORMING PARTICLES," and / or any suitable cartridge may be used.

[0014] The cartridge preferably has a single z-plane through which the particles flow, however, the cartridge may have more than one z-plane (e.g., depending on the flow path, flow rate, temperature, Reynolds number, etc.).

[0015] The particle velocity is preferably around 4 m / s (e.g., 1-6 m / s, 2-4 m / s, 3-5 m / s, 3-4.5 m / s, 1-4 m / s, 3 m / s, 3.5-4.5 m / s, 3.6 m / s, 3.7 m / s, 3.8 m / s, 3.9 m / s, 4 m / s, 4.1 m / s, 4.2 m / s, 4.3 m / s, 4.4 m / s, 5 m / s, values ​​or ranges therebetween, etc.), but may be less than about 1 m / s and / or greater than about 6 m / s. In some variations, the accuracy and / or reliability of the results (e.g., biophysical parameters, indices, health status, etc.) may depend on the particle velocity. For example, when the flow rate is not around 4 m / s, the results may be difficult to tune, inaccurate, require non-linear adjustments, have chaotic adjustments, require greater adjustments, and / or otherwise have lower reliability and / or precision (e.g., compared to cases with particle velocities of around 4 m / s), but may still be used (e.g., to determine health conditions with lower precision, reliability, repeatability, etc.). However, the reliability may be independent of particle velocity (e.g., different models may be trained, used, etc. with different particle velocities to ensure accuracy, reliability, etc.), and / or the accuracy and / or reliability may be differently dependent on particle velocity.

[0016] The microfluidic device can include and / or be coupled to a flow device that functions to generate a fluid flow within the microfluidic device. The flow device can include a pump, an impeller, and / or any other flow device. The flow device can provide negative and / or positive pressure (e.g., pushing, pulling, etc., fluid through the microfluidic channel).

[0017] The detection system is preferably an image acquisition system (200), such as a microscope (e.g., an inverted microscope, an upright microscope, etc., configured to acquire bright field images, dark field images, fluorescent images, etc., of the sample), a camera, and / or any image acquisition system. Images are preferably acquired at a frame rate that is at least 10,000 frames / second, such as 20,000, 50,000, 100,000, 200,000, 300,000, 500,000, 750,000, 1,000,000, 2,000,000, 5,000,000, 10,000,000 frames / second. However, images can be acquired at less than 10,000 frames / second and / or at any rate. The field of view (250) of the image acquisition system is preferably centered proximate to a stagnation point of the deformation region (e.g., a flow path and / or a point of force equilibrium). However, the stagnation point may be an edge, a corner, arbitrarily, at a predetermined location, randomly, and / or otherwise aligned with respect to the image acquisition system (e.g., within the captured image). However, the image acquisition system may be otherwise aligned.

[0018] The computing system (300) may function to process (e.g., analyze) the data set, control system operations (e.g., trigger data acquisition, time data acquisition, modify flow rates, etc.), determine biophysical parameters, determine health scores, adjust biophysical parameters (and / or health scores, indices, metrics, etc.), determine the health status of the patient, and / or otherwise function. The computing system may be local (e.g., a dedicated computer co-located with the data acquisition system), remote (e.g., a cloud computer, server, database, etc.), and / or distributed in any manner. The computing system comprises one or more of a processor, microprocessor, computer processing unit, graphics processing unit, and / or any suitable processor. In a particular example, the computing system may comprise GPU-accelerated computation, which may be beneficial for quickly and / or efficiently handling (e.g., processing) large data sets (such as those generated by image sets).

[0019] The computing system may include an image analyzer (320), a filter (340), a conditioner (360), an index calculator, a health scorer (380), and / or any other suitable components or modules. The modules may run on the same or different processors, using different or the same executable files (e.g., programs), and / or may execute or operate differently.

[0020] The image analyzer (320) preferably functions to analyze a set of images. For example, the image analyzer may identify features in the images, process the images (e.g., apply one or more image corrections, warp the images, etc.), and / or otherwise function. The image analyzer may comprise a machine learning algorithm (e.g., K-means clustering, mean shift clustering, density-based spatial clustering for applications with noise, expectation maximization (EM) clustering with Gaussian mixture models, agglomerative hierarchical clustering, etc.), a neural network (e.g., convolutional neural network, recurrent neural network, deep neural network, etc.), a ray tracer, a classifier (323) (e.g., a binary classifier), and / or may use any suitable equations and / or algorithms. The image analyzer may include an edge detector, an object detector, and / or any suitable detector.

[0021] In some variations, the image analyzer can classify objects in the sample. For example, white blood cells can be classified (e.g., class, probability that the particle is in the class, etc.) into a particular white blood cell subpopulation (e.g., monocytes, lymphocytes, neutrophils, basophils, eosinophils, etc.). In another example, particles can be classified as white blood cells or not white blood cells (e.g., members of a class, probability that the particle is in the class, etc.). However, objects can additionally or alternatively be classified as red blood cells, reticulocytes, platelets, cellular debris, mechanical debris, dust, and / or another suitable classification. Objects can be classified for each image, for each event (e.g., multiple images of a given particle or object), for populations (e.g., where the sample is sorted into a given cell subpopulation, preferably with at least a threshold confidence of sorting), and / or otherwise classified.

[0022] The image analyzer can include an occlusion detector (327), which can function to determine when two (or more) particles are touching and / or co-located within an image and / or event. For example, the occlusion detector can detect occlusion based on particle size, particle dynamics, light intensity, image blur, based on state history (e.g., particle trajectories, changes between frames, etc.), and / or in any manner.

[0023] The image analyzer may include a parameter extractor, which preferably functions to determine biophysical parameters from a data set (e.g., an image set, an analyzed image, etc.). The parameter extractor may use equations, lookup tables, neural networks (e.g., convolutional neural networks, recurrent neural networks, deep neural networks, etc.), machine learning, and / or any suitable algorithm to determine the biophysical parameters. The biophysical parameters may include frame parameters (e.g., parameters extracted from a single image, parameters associated with a single image, etc.), event parameters (e.g., parameters extracted from multiple images of a common particle, parameters associated with multiple images of a common particle, etc.), population parameters (e.g., parameters extracted from images of multiple particles, distribution of frame parameters and / or event parameters of multiple particles, etc.), and / or any suitable parameters.

[0024] Exemplary biophysical parameters may include structural parameters, trajectory parameters, patient parameters, positional parameters (e.g., cell position relative to an image frame, cell position relative to a channel, cell position relative to an outlet, cell position relative to an inlet, cell position relative to a stagnation point, etc.), and / or any parameter. Examples of structural parameters include shape (e.g., ellipticity, helicity, oblongness, circularity, curvature, skewness, etc.), aspect ratio (e.g., ratio of longest dimension to shortest dimension, ratio of length to width, etc.), size (e.g., lateral extent, longitudinal extent, depth, height, width, length, volume, surface area, etc.), constituent structure (e.g., location of cell membrane, shape of cell, structure of cell wall, etc.), constituent morphology (e.g., cell morphology, cell shape, particle shape, etc.), internal structure (e.g., shape, form, size of nucleus and / or other organelles of a cell, etc.), and / or any other structural parameter. Examples of patient parameters include complete particle counts, complete blood counts, complete white blood cell counts, complete neutrophil counts, complete monocyte counts, complete lymphocyte counts, complete basophil counts, complete eosinophil counts, complete red blood cell counts, complete platelet counts, component counts (e.g., number of cells and / or cell segments, number of particles and / or particle segments, etc.), presenting symptoms (e.g., patient temperature, blood pressure, weight, blood oxygenation, etc.), cell density, cell culture results, hydration, and / or any other sample parameters.

[0025] A trajectory parameter can be a parameter related to and / or determined from the trajectory of a cell through a microfluidic cartridge (e.g., a deformation region, a focusing region, etc. of a microfluidic cartridge). A trajectory can be a series of discrete positions of a cell (e.g., a center of gravity, a center of mass, a reference point, an average position of a cell, etc.), a continuous path of a cell, and / or corresponds to any motion of a cell as it passes through a fluid flow (e.g., a deformation region of a microfluidic cartridge). Examples of trajectory parameters include a direction of motion of an object and / or feature, a velocity of motion of an object and / or feature (e.g., an average velocity, an instantaneous velocity, etc.), an acceleration of the motion of an object and / or feature, a vibration of the motion of an object and / or feature (e.g., an amplitude of vibration, a frequency of vibration, a phase of vibration, a modulation of vibration, a damping of vibration, etc., as shown in the example of FIG. 9), a viscoelastic inertial response (VEIR), a deviation in a particle flow trajectory (e.g., from a linear path, an expected path, etc.), and / or any other trajectory parameter. The trajectory parameters can be determined based on differences, sums, amplitudes, maxima, minima, averages, and / or other characteristics of one or more positions of the trajectories. In variations in which the trajectory parameters include vibrations, the vibrations may correspond to vibrations of a reference point (e.g., a centroid, an extremum, a medoid, etc.), a reference axis (e.g., one or more dimensions such as length, width, depth, etc.), one or more reference surfaces (e.g., an object boundary, an internal boundary of the object such as corresponding to an organelle boundary, etc.), a reference volume (e.g., an object volume, an internal structure of the object, etc.), and / or other parts of the object. The vibrations preferably occur along a reference axis perpendicular to the direction of motion of the object. However, the reference axis may be parallel to the direction of motion and / or have any orientation relative to the direction of motion. The vibration amplitude is preferably on a micron size scale (e.g., 1-10 μm, 10-100 μm, etc.), but may additionally or alternatively be on a nanometer scale (e.g., 1-100 nm, 100 nm to 1 μm, etc.), and / or any suitable distance.

[0026] The adjuster (360) functions to adjust the metrics (e.g., scores, indices, biophysical parameters, etc.). The metrics are preferably adjusted based on system characteristics and / or particle characteristics, but may be adjusted based on image characteristics, event characteristics, population characteristics, and / or based on any suitable information. Exemplary system characteristics include flow rate, temperature (average temperature, instantaneous temperature, etc.), temperature fluctuation (e.g., across measurements, specified fluctuation, etc.), flow rate fluctuation (e.g., across measurements, specified fluctuation, etc.), pressure, pressure fluctuation (e.g., across measurements, specified fluctuation, etc.), humidity, channel size, and / or any suitable system characteristic. Exemplary particle characteristics include handedness (e.g., direction of rotation of fluid flow, direction of particle propagation, etc.), particle population (e.g., white blood cell subpopulation), particle size, particle shape, particle trajectory, particle acceleration, particle planarity, and / or any suitable particle characteristic. However, the metrics may be adjusted in any manner.

[0027] The adjuster preferably applies a linear adjustment, but may apply a non-linear adjustment (e.g., quadratic, cubic, quartic, exponential, logarithmic, hyperexponential, sinusoidal, hyperbolic, hyperbolic trigonometric, inverse trigonometric, inverse polynomial, etc.), and / or any suitable adjustment. In a first specific example (e.g., as shown in FIG. 4), the adjustment is m 調整済み =m 判定済み +k*(Q 0 -Q 測定済み ), where m 調整済み may be an adjusted metric, m 判定済み may be a determined (e.g., measured) metric, k may be a constant (e.g., one that may be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc.), and Q 0 can be a reference characteristic value (e.g., a reference flow rate such as a target value, average value, mean value, mode, median value, set value, or an expansion value such as a Taylor expansion), and Q 測定済み may be a measured characteristic value. In a second specific example (e.g., as shown in FIG. 6), the adjustment is 調整済み =m 判定済み+c*(ξ), where c can be a constant (e.g., one that can be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc.) and ξ can be the handedness of the fluid flow (e.g., the handedness of the particle trajectories). In a third example (e.g., as shown in FIG. 8), the adjustment is m 調整済み =m 判定済み +m*(z-plane), where m can be a constant (e.g., one that can be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc.), and z-plane can be the z-plane of the particle (e.g., absolute z-height within the channel, z-height relative to a reference for the channel, z-height relative to a second z-plane of the channel, etc.). In a fourth specific example, the adjustment is m 調整済み =m 判定済み +k*(Q 0 -Q 測定済み )+c*(ξ)+m*(z-plane), however, any suitable adjustment can be applied.

[0028] The regulator is preferably used only when the system characteristic is within a threshold range of the setpoint (e.g., average, target, setpoint, etc.). However, the regulator can be used when the system characteristic is outside the threshold range of the setpoint. This can be beneficial to ensure that the parameter is in a linear regime and can be corrected using linear correction (which is often more reliable, more repeatable, less contaminated by cross-correlation with other parameters, and / or may otherwise be desirable or provide technical advantages). The threshold range can be symmetrical and / or asymmetrical. The threshold range can be absolute, relative, and / or any suitable value. For example, the threshold range can be ±0.1%, ±0.5%, ±1%, ±2%, ±5%, ±10%, values ​​or ranges therebetween, and / or any suitable range of values ​​about the target value. In an illustrative example, the regulator can be used when the system flow rate is within (above and / or below) 4% of the target flow rate.

[0029] If the system characteristic falls outside the threshold range, the data is typically rejected (e.g., another sample is collected, measured, analyzed, etc.), but the data can additionally or alternatively be adjusted (e.g., using a non-linear adjustment, using a linear adjustment, etc.), flagged (e.g., for potential inaccuracies), used without adjustment, and / or otherwise processed and / or used.

[0030] In one example, the adjuster can adjust a metric associated with neutrophils. In a second example, the adjuster can adjust a metric associated with neutrophils having a particular handedness. In a third example, the adjuster can adjust a metric associated with neutrophils having a particular handedness in one way and adjust a metric associated with neutrophils having the other handedness in a different way. In a fourth example, the adjuster can adjust a metric associated with neutrophils having a particular handedness in one way and adjust a metric associated with monocytes in a different way. However, the adjuster can adjust the metrics in any way.

[0031] The index calculator preferably functions to determine an index from the biophysical characteristics. Each particle (e.g., each particle or object to be included in the health score determination) may also be associated with an index (e.g., an event index). However, the index may be a global index (e.g., a single index associated with a patient), a subpopulation index, an image index, and / or any suitable index. The index calculator may include machine learning algorithms, neural networks, equations, lookup tables, heuristics, and / or any suitable algorithm. In an illustrative example, the index calculator may incorporate particle size, particle trajectory parameters, cell count, cell count of single cells or white blood cell populations, and / or may determine the index with any suitable input.

[0032] The health scorer preferably functions to determine the health status of the patient. The health scorer (380) may be the same as the index calculator and / or may be different from the index calculator. The health scorer may include machine learning algorithms, neural networks, equations, look-up tables, heuristics, and / or any suitable algorithm. The health status is preferably determined based on the index, but may additionally or alternatively be determined based on biophysical characteristics, images, data sets, and / or any suitable information. In some variations, auxiliary information (such as presentation of presenting symptoms, patient parameters, sex, race, age, height, weight, body mass index, time of day, recent travel, recent exposure, etc.) may be used (in addition to and / or as an alternative to the index) to determine the health status of the patient. The health status may be a diagnosis, a probability of a diagnosis, a scaled value (e.g., severity of the condition, degree of activation, etc.), and / or any suitable value. For example, the health condition may indicate a patient's immune activation state and / or degree of immune activation (e.g., cancer, metastatic cancer, sepsis, COVID-19 infection, sepsis-related disease, anemia, bleeding conditions, systemic inflammatory response syndrome, blood cell health, organ health, inflammation, cytokine release, cytokine storm, autoimmune disorders such as arthritis, rheumatic disorders such as lupus, graft versus host disease, etc.). In some variations, the health condition may indicate a cause of the health condition (e.g., bacteria, fungus, virus, chemical, etc.) and / or any suitable information.

[0033] 4. Method. The method is preferably operable to determine a health state of a patient, but may additionally or alternatively be operable to determine a biophysical parameter (and / or a biophysical parameter distribution) of the patient, and / or perform any suitable function. The method includes receiving a dataset (S100), optionally filtering the dataset (S200), determining a metric (S300), optionally adjusting the metric (S400), determining a health state of the patient (S500), optionally providing an intervention to the patient (S600), and / or any suitable function. The method may be performed using the above-mentioned system and / or any suitable system, such as the system disclosed in U.S. patent application Ser. No. 17 / 575,388, entitled "SYSTEM AND METHOD FOR DETERMINING AN IMMUNE ACTIVATION STATE," filed January 13, 2022, and / or U.S. patent application Ser. No. 17 / 575,422, entitled "SYSTEM AND METHOD FOR DETERMINING AN IMMUNE ACTIVATION STATE," filed January 13, 2022, each of which is incorporated by reference in its entirety herein.

[0034] The method (and / or steps thereof) can be performed in parallel (e.g., concurrently, contemporaneously, simultaneously, etc.) and / or sequentially (e.g., consecutively) on different samples (e.g., samples associated with different patients). The steps of the method can be performed contemporaneously (e.g., simultaneously, concurrently, etc.). For example, images associated with one event can be processed while another event is being measured (e.g., (S200-S400) can be performed on one cell while another cell is being measured). However, the steps of the method can be performed sequentially and / or in any suitable order or manner.

[0035] The step of receiving a data set (S100) functions to receive a data set associated with a patient. The data set can be received from a cytometer (e.g., a flow cytometer, a deformability cytometer, etc.), an image acquisition system (e.g., a camera), a computing system (e.g., a memory, a database, a storage device, etc.), and / or from any suitable data source. The data set is preferably a plurality of images of the patient's blood sample (e.g., a plurality of images of each cell of the patient's blood sample), but may additionally or alternatively include electrical data (e.g., resistance, conductivity, etc.), thermal data, probe measurements (e.g., a probe in contact with the sample, a probe in contact with a microfluidic channel, a probe in contact with a sheath fluid, etc.), pressure measurements, force measurements, and / or any suitable data.

[0036] Receiving the dataset may include processing the dataset. The dataset may be processed in real-time or near real-time (e.g., as the dataset is acquired), delayed (e.g., until the entire dataset is acquired, until a target time for processing, etc.), and / or processed at any suitable time. Processing the dataset may include segmenting the dataset (e.g., divided into events, into sets of frames associated with shared particles, into discrete time series, etc.), classifying the dataset (e.g., using a binary classifier, using a multi-class classifier, e.g., into particles or non-particles, into specific particle populations, etc., as shown in FIG. 7A), image transforming (e.g., warping, scaling, translating, rotating, etc., color balancing, etc.), and / or otherwise processing the dataset or portions thereof. In some variations, a processed dataset may be received (e.g., when the processed data is stored and retrieved for later analysis).

[0037] By way of illustration, a dataset can be segmented into events, where each event can include a time series of frames (e.g., images) that include a common particle. Two or more events can share images (e.g., an image can include two or more particles). However, images can be unique to an event, and / or events and images can be otherwise related or associated with each other. Images can be segmented using neural networks, using machine learning, manually, based on time (e.g., a segment can include a predetermined number of frames, frames acquired within a predetermined time of the first frame that includes a particle, etc.), and / or otherwise.

[0038] In another example, the events of the data set (or particles thereof) can be classified into particle subpopulations. For example, the particles can be classified into white blood cells, white blood cell subpopulations, red blood cells, reticulocytes, platelets, cell debris, mechanical debris, dust, and / or any suitable classification, and the events can be processed (e.g., analyzed, conditioned, etc.) based on the classification. The events can be classified using neural networks, machine learning, based on image characteristics (e.g., features detected, number of features, shape of features, light detected, etc.), and / or in any manner. As a specific example, the particles can be classified into subpopulations based on cell size, optical intensity, organelle size, number of visible organelles (e.g., number of nuclei), and / or any suitable information.

[0039] The step of filtering the dataset (S200) preferably functions to select data (e.g., high quality data, data having sufficient quality, data that meets one or more criteria, etc.) from the dataset to be analyzed (e.g., processed dataset, acquired dataset, raw dataset, measured dataset, etc.). The dataset may be filtered before, during, and / or after the dataset is received. The dataset is preferably filtered using the computing system (e.g., its filters), but may be filtered using filters of the measurement module (e.g., cytometer) and / or using any suitable component. The data sets can be filtered based on measurement system characteristics (e.g., flow rate, temperature, humidity, pressure, etc.), patient sample characteristics (e.g., cell size, cell shape, cell count, biophysical properties, ancillary data, etc.), image characteristics (e.g., brightness, contrast, occlusion, z-plane, type of particle, etc.), event characteristics (e.g., particle trajectory, particle inlet position (252), particle inlet velocity, particle exit position (258), particle exit velocity, particle exit direction, particle rotation, etc.), population characteristics, and / or based on any suitable characteristics. The data (and / or data sets) can be filtered before receiving the data set, during reception (e.g., data associated with a first cell can be filtered while data for a second cell is being received), and / or after reception.

[0040] The filtered data set preferably includes data from about 10,000-100,000 particles (e.g., patient cells), but may include data for fewer than 10,000 particles and / or more than 100,000 particles. For each particle, the data preferably includes at least 10 frames (e.g., images, consecutive frames, etc.) that include the particle, which may provide technical advantages, particularly when determining trajectory parameters of the particle. However, fewer than 10 frames may be sufficient (e.g., for particle counting, for morphological analysis, etc.). A set of frames associated with (e.g., including) a particle may be referred to (e.g., collectively) as an event and / or may be referred to in any manner.

[0041] The dataset (and / or the data therein) may be filtered using a tree filter, a combinatorial filter (e.g., filtering based on the filter scores of the data for different filters), a cutoff filter (e.g., high pass, low pass, band pass, band stop, notch, etc.), a Bloom filter, and / or any suitable filter structure may be used.

[0042] Datasets can be filtered at a population level (e.g., to reject and / or accept datasets for analysis), at an event level (e.g., to reject and / or accept events in a filtered dataset), at a frame level (e.g., to reject and / or accept frames for inclusion in an event, to reject and / or accept frames for inclusion in a dataset, etc.), and / or at any suitable level.

[0043] Generally, if the data and / or dataset does not meet a threshold (e.g., a filter value), the data (and / or dataset) is excluded from the filtered dataset (e.g., not analyzed). However, the data may additionally or alternatively be flagged (e.g., for having a value outside the threshold), adjusted (e.g., scaled, transformed, etc. to meet the threshold), and / or otherwise manipulated and / or modified.

[0044] Examples of population level filtering (e.g., population level thresholds) can include particle size distribution (e.g., distribution shape, distribution variability, etc.), system characteristics (e.g., flow rate, temperature, humidity, pressure, atmospheric pressure, etc.), particle subpopulation distribution (e.g., relative numbers of different leukocyte subpopulations, number of identified leukocyte subpopulations, confidence in leukocyte identification, etc.), and / or any suitable population level filter can be applied.

[0045] Examples of event level filters (e.g., event level thresholds) can include occlusion filters, particle velocity filters, particle time filters (e.g., the amount of time a particle spends in a deformation region), particle trajectory filters, particle entry filters, particle exit filters, particle z-plane, particle subpopulations, and / or any suitable filters.

[0046] Examples of frame level filters (e.g., frame level thresholds) may include an occlusion filter, a blur filter, a particle count filter (e.g., the number of particles present in a frame), a color filter, a brightness filter, a contrast filter, a focal plane filter (e.g., a sharpness filter), and / or any suitable filter.

[0047] In a first specific example (e.g., as shown in Figures 5A, 5B, and / or 5C), the filtered data set can include data (e.g., events) associated with particles entering and / or exiting a region of interest of the image within a threshold window (e.g., at least a threshold distance from the edge of the channel, near the center of the channel, an area between approximately 25-75% of the opening or outlet of the channel, within a threshold pixel window of the image field of view, etc.) and can exclude data associated with particles entering and / or exiting the region of interest outside the threshold window.

[0048] In a second specific example (e.g., as shown in FIG. 7A and / or FIG. 7B), the filtered data set can include data (e.g., events, images, etc.) related to isolated particles (those that are not obscured, such as with other particles) and can exclude data related to obscured particles. Obscured particles can be identified based on particle size (e.g., particularly, but not limited to, before the particles are deformed), based on particle number, based on change in particle position between frames, based on image intensity (e.g., contrast of particle compared to expected contrast), and / or in any manner. In a variation of the second specific example, frames and / or events that include two or more particles can be included in the filtered data set and / or otherwise included in and / or excluded from the filtered data set, particularly, but not limited to, if the two or more particles are isolated from each other, are present only in a subset of images, have a threshold distance between the particles, have a threshold timing difference between the particles (e.g., a threshold time difference between particle entrances).

[0049] In a third specific example, events can be filtered based on the z-plane in which the particle is traversing the channel. For example, particles detected in the z-plane of the channel can be included in the filtered data set, and particles detected in a different z-plane can be excluded from the filtered data set. The included z-plane can be a preferred z-plane, a z-plane in the focal plane of the image acquisition system, a first z-plane (e.g., a z-plane in which the first measurement particle is present, a z-plane selected from a previous sample, etc.), a z-plane at least a predetermined distance from the wall of the channel (e.g., bottom, top, etc.), and / or any suitable z-plane can be selected.

[0050] In a fourth specific example, events may be filtered based on the number of consecutive frames associated with the event, e.g., if more than 10 frames are associated with a particle, the data may be included in the filtered data set, and / or if less than 10 frames are associated with a particle, the data may be excluded from the filtered data set.

[0051] In a fifth specific example, the data set can be filtered based on the flow rate in the cartridge. For example, data can be included in the data set if the flow rate is within a threshold value (e.g., about ±0.5%, 1%, 2%, 5%, 10%, 20%, values ​​or ranges therebetween, etc.) of a target flow rate.

[0052] In a sixth specific example, the data set may be filtered based on the temperature of the sample being measured. For example, if the temperature is within a threshold of the target temperature (e.g., about ±0.5%, 1%, 2%, 5%, 10%, 20%, values ​​or ranges therebetween, ±0.1°, ±0.5°, ±1°, ±2°, ±5°, ±10°, values ​​or ranges therebetween, etc.), the data may be included in the data set.

[0053] In a seventh specific example, the dataset may be filtered based on particle types associated with the data. For example, filtering the dataset may include determining particle types and determining whether to include (or exclude) those associated with the particles in the dataset based on particle types. Particle (e.g., cell) types may be determined based on particle color, nucleus size, nucleus shape, number of nuclei, granule color (e.g., stained granules), granule location relative to the nucleus, particle shape, and / or in any suitable manner. Illustratively, the dataset may include images and / or data derived therefrom (e.g., metrics, biophysical parameters, etc.) associated with neutrophils and / or monocytes (neutrophils and monocytes may be distinguished based on, for example, size, nucleus shape, nucleus size, number of organelles, etc.).

[0054] In an eighth example, any or all of the first through seventh example filtering examples above may be combined and / or applied. Filter results may be combined by voting (e.g., voting to include or exclude data from the dataset), preference (e.g., ranked preferences for filtering based on filter results), absolute filtering (e.g., where data is excluded by any filter that excludes data from the dataset), a threshold number of filters may be used to include and / or exclude data (e.g., filters of 1, 2, 3, 4, 5, 6, 7, 10, etc.), and / or filters may be combined otherwise.

[0055] However, the data and / or datasets can be filtered in any manner.

[0056] The determining metric step (S300) serves to determine a metric (e.g., a biophysical parameter, etc.) associated with each particle in the dataset (e.g., the processed dataset, the filtered dataset, the measured dataset, etc.). The metric is preferably determined using a computing system (e.g., an image analyzer, etc.), but may be determined using any suitable system and / or component. The metric may be used (directly, indirectly, etc.) to determine the health status of the patient. The metric may be numerical, distributional, categorical, normalized, relative, absolute, and / or any suitable value. The metric is preferably determined after filtering the dataset, but may be determined before and / or during dataset filtering (e.g., a metric for a first cell may be determined while data for a second cell is being filtered).

[0057] The metrics may be determined using machine learning algorithms, neural networks, equations, heuristics, fits, estimators (e.g., Monte Carlo simulation, Kalman filters, extended Kalman filters, etc.), and / or using any suitable algorithms and / or methods.

[0058] The inputs used to determine the metrics may include a dataset (e.g., a filtered dataset, a processed dataset, a received dataset, etc.), a subset of a dataset (e.g., event data, frame data, etc.), system data (e.g., cartridge shape, cartridge size, channel size, channel shape, temperature, flow rate, pressure, etc.), auxiliary data (e.g., patient data, patient demographics, etc.), and / or any suitable information. The metrics (e.g., outputs) may include biophysical properties (e.g., of particles, samples, etc.), indices (e.g., white blood cell structure index), and / or any suitable information. In some variations, one or more metrics may be used to determine other metrics.

[0059] Determining the metric may include determining one or more system characteristics (S350). The system characteristics may be determined from the data set (or a subset thereof), from sensor data, and / or from any suitable data. In an illustrative example, the flow rate may be determined using multiple images of one or more particles in the channel (e.g., in a deformation region of the channel, in a focused region, in a linear region, in a region that does not divert particles, in a flow measurement region, etc.). In this example, the flow rate may be determined based on the frame rate (e.g., the amount of time between images acquired), particle displacement (e.g., the distance a particle travels between images), particle size (e.g., average particle size, expected particle size, etc.), temperature, Reynolds number, channel aspect ratio, channel geometry, Dean number, channel hydraulic diameter, particle confinement ratio, and / or using any suitable information. In a variation of an exemplary embodiment, the step of determining the flow rate (S355) segments the multiple images into a foreground (e.g., one or more cells or particles) and a background, determines the displacement of the foreground between images, and determines the flow rate based on the displacement (in addition to image frame rate, particle size, channel size, etc.).

[0060] In some variations, the system characteristic may be modified (S370) after determining the metric (e.g., modify the system characteristic based on the metric). By way of illustration, the flow rate may be corrected based on the measured particle size distribution.

[0061] Determining the metrics can include aggregating the metrics, which can serve to summarize the metrics, determine sample-level characteristics, and / or otherwise function. The aggregated metrics can be referred to as features and / or by any suitable name. Metrics can be aggregated at the sample level, at the sample subpopulation level, based on shared characteristics (e.g., can be aggregated based on shared z-planes, shared handedness, etc.). In some variations, the metrics can be adjusted (e.g., as in (S400)) prior to metric aggregation. Metrics can be aggregated based on particle subpopulations, one or more statistical descriptors, and / or based on any suitable information. Examples of statistical descriptors include the mean, n-th percentile values ​​(e.g., percentile values ​​of 0.1, 1, 5, 10, 20, 25, 50, 75, 80, 90, 95, 99, 99.9, values ​​therebetween, etc.), distribution moments (e.g., skewness, kurtosis, L-moment, etc.), variance (e.g., variance, standard deviation, coefficient of variation, range, interquartile range, etc.), mode, median, dispersion index, and / or any suitable statistical descriptor. Some examples of aggregated metrics include mean particle size (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), mean VEIR (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), 75th percentile VEIR (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), mean aspergillus size ... In one embodiment, the metric may be determined based on the total number of neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, aspect ratio (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), aspect ratio variance (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), VEIR variance (e.g., for neutrophils, monocytes, lymphocytes, neutrophils, basophils, eosinophils, combinations or partial combinations thereof, etc.), and / or any suitable aggregation metric.Each aggregated metric may be a scalar (e.g., a single value), a vector (e.g., a set of values ​​for each subpopulation, a set of values ​​for the aggregated metric, etc.), a matrix (e.g., including correlations between the aggregated metrics), a tensor, and / or have any suitable form.

[0062] The metric adjusting step (S400) preferably functions to modify, transform, normalize, process (e.g., pre-process, post-process, etc.), weight, re-specify variables, scale, filter, and / or otherwise modify the metric. The metric adjusting step may provide technical advantages of increasing reproducibility, reliability, repeatability, and / or otherwise improving the patient metric and / or health status determination. The metric is preferably adjusted using a computing system (e.g., a adjuster, etc.), but may be determined using any suitable system and / or component. The metric may be adjusted before (S300) (e.g., a particle-related metric may be adjusted before the feature is determined), during (S300) (e.g., a first particle-related metric may be adjusted while a second particle-related metric is determined), and / or after (S300). The adjusted metric may be a determined metric, an aggregated metric (e.g., a feature), an index (e.g., a value derived from a metric, an aggregated metric, etc., such as in (S500)), and / or any suitable metric.

[0063] The metrics are preferably adjusted linearly, but may be adjusted nonlinearly. The metrics are preferably adjusted in the same way, but may be adjusted differently. The adjustments (e.g., adjustment coefficients) may depend on system characteristics (e.g., flow rate, temperature, humidity, pressure, channel geometry, etc.), sample characteristics (e.g., sample concentration, sample composition, etc.), particle characteristics (e.g., particle size, particle metric, particle population, particle type, etc.), and / or any suitable information. The adjustments (e.g., adjustment coefficients) may be empirically, heuristically, physically determined, determined using machine learning, determined using neural networks, and / or determined in other ways. In a first specific example, the adjustment coefficients may depend on the particle subpopulation. For example, a first adjustment may be used to adjust a metric associated with neutrophils, and a second adjustment may be used to adjust a metric associated with monocytes. In a second specific example, which adjustment is applied may depend on the particle subpopulation. For example, a z-plane adjustment may be applied to neutrophils (as an example, lymphocytes are not included). However, any adjustment may be applied to any subpopulation, and / or any suitable adjustment may be applied to any suitable particle.

[0064] In a first specific example (e.g., as shown in FIG. 4), the adjustment is m 調整済み =m 判定済み +k*(Q 0 -Q 測定済み ), wherein m 調整済み can be an adjusted metric, m 判定済み may be a determined (e.g., measured) metric, k may be a constant (e.g., something that may be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc., an adjustment factor, etc.), and Q 0 can be a reference characteristic value (e.g., a reference flow rate such as a target value, average value, mean value, mode, median value, set value, or an expansion value such as a Taylor expansion), and Q 測定済み can be the measured characteristic value. In a second specific example (e.g., as shown in FIG. 6), the adjustment is 調整済み =m 判定済み+c*(ξ), where c can be a constant (e.g., one that can be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc.) and ξ can be the handedness of the particle (e.g., the handedness of the particle trajectory). In a third example (e.g., as shown in FIG. 8), the adjustment is m 調整済み =m 判定済み +m*(z-plane), where m can be a constant (e.g., one that can be empirically determined, measured, derived, a physical constant, a relationship, a unit conversion, etc.) and z-plane can be the z-plane of the particle (e.g., absolute z-height within the channel, z-height relative to a reference for the channel, z-height relative to a second z-plane of the channel, etc.). In a fourth specific example, the adjustment includes m 調整済み =m 判定済み +k*(Q 0 -Q 測定済み )+c*(ξ)+m*(z-plane). However, any suitable adjustment may be applied.

[0065] In some variations, (S400) may include determining a particle type, and the adjustment may depend on the particle type. The type of particle (e.g., cell) may be determined based on (e.g., by determining, identifying the particle based thereon, etc.) the color of the particle, the size of the nucleus, the shape of the nucleus, the number of segments, the color of the granules (e.g., stained granules), the position of the granules relative to the nucleus, the shape of the particle, and / or any suitable manner. By way of illustration, if the cell type of the white blood cell is a neutrophil, adjusting (linearly) the one or more biophysical parameters further includes adjusting the one or more biophysical parameters based on the z-plane of the white blood cell. However, the metric(s) may be adjusted in any manner based on the particle type.

[0066] In some variations, (S400) may include adjusting system characteristics. For example, the system flow rate may be modified so that the measured flow rate matches a target flow rate (thereby eliminating the need to adjust the flow rate). In another example, the temperature of the sample (and / or cartridge) may be modified to ensure that the sample is within a threshold range of a target temperature during measurement. The system characteristics are preferably adjusted prior to measuring the data set, but may be adjusted during measurement.

[0067] The step of determining health status (S500) preferably functions to determine the patient's immune activation and / or response. The health status is preferably determined using a computing system (e.g., an index calculator, health scorer, etc.), but may be determined using any suitable system and / or component. The patient's health status may include an index (e.g., a white blood cell structure index, a normalized index such as 0-1, a non-normalized index such as 0-10, 0-100, 0.1-10, 0.1-100, 1-10, 1-100, etc.), a diagnosis, a classification, a probability of a diagnosis applicable to the patient, and / or any suitable information.

[0068] The health status may be determined from a model, an equation, a lookup table, a machine learning algorithm (e.g., trained to output a score, trained to output an index, trained to output a disease status, etc.), and / or may be determined otherwise. Examples of models that may be used include logistic regression (e.g., linear logistic regression, non-linear logistic regression, etc.), decision trees, Bayesian classifiers, nearest neighbor techniques, support vector machines, decision forests (e.g., random forests), neural networks, gradient boosting, and / or any model may be used. The model may be a predefined model, a general model, a model selected based on auxiliary information, and / or any other suitable model.

[0069] Inputs to the model may include one or more metrics (e.g., determined metrics, adjusted metrics, aggregated metrics, features, adjusted and aggregated metrics, adjusted features, etc.), one or more parameters, auxiliary information such as age, maturity (e.g., infant, toddler, child, adolescent, adult, etc.), weight, height, race, sex, temperature, body mass index, percentage of body fat, pre-existing diseases, duration of symptoms, onset of symptoms, migration, blood oxygenation level, blood pressure, source of infection, treatment parameters (e.g., hospitalization, treatment administered, etc.), collection parameters (e.g., temperature, other compounds found in the sample, etc.), images (e.g., from a set of images), images of cells and / or analyses thereof under different lighting conditions (e.g., different intensities of light, different wavelengths of light, etc.), different event-level information, frame-level analysis, population analysis, and / or any input. In an illustrative example, the inputs may include aggregated trajectory parameters (e.g., mean VEIR, 75th percentile VEIR, mean VEIR for monocytes, 75th percentile VEIR for monocytes, etc.), cell size (e.g., mean cell size, mean cell size for neutrophils, etc.), and / or cell counts (e.g., number of neutrophils, number of white blood cells, etc.). The inputs are preferably determined based on a training dataset, but may be determined in any manner.

[0070] The output of the model may include an index (e.g., a health status index, a Leukocyte Structure Index (LSI), an IntelliSep Index (ISI), a value between 0-1, 0-10, etc., a score, a number of components having parameters and / or combinations of parameters above a threshold, such as a total number of components above a threshold, a percentage of the total number of components above a threshold, a probability, etc.), a health state, a severity of the health state, a probability of the health state, and / or any output.

[0071] In an illustrative example, the health status may be determined based on the index: If the index exceeds a threshold (e.g., 0.01, 0.05, 0.1, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.75, 0.8, 0.9, 0.95, 0.99, 1, etc.), the health status may indicate that the patient has the disease (or may have the disease with at least a 50% probability, etc.). If the index is below a threshold (e.g., 0.01, 0.05, 0.1, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.75, 0.8, 0.9, 0.85, 0.99, 1, 2, 3, 4, 5, 6, 7, 8, 9, etc.), the health status may indicate that the patient does not have the disease (or is likely not to have the disease, such as at least a 50% chance that the patient does not have the disease). However, an index below the threshold may indicate disease, an index above the threshold may indicate no disease, and / or the index may be used in any manner. The threshold may be determined based on a training dataset (e.g., in a manner similar to how the model is generated, in a manner similar to how the inputs to the model are selected), pre-determined, determined according to an equation (e.g., an equation based on and / or accounting for the auxiliary sample information), and / or determined in other ways.

[0072] Providing an intervention (S600) may function to provide an intervention to the patient. The intervention is preferably dependent on a health state (e.g., immune activation state), but may be dependent on biophysical characteristics, patient data, and / or any suitable data and / or information. Providing an intervention is preferably performed automatically (e.g., a computing system generates a suggestion based on the health state), but may be performed manually and / or in any manner. Examples of interventions include tracking the progression of a patient's disease state over time, determining treatment parameters (e.g., estimated length of hospital stay, estimated triage requirements, etc.), initiating secondary analyses (e.g., innate immune activation assays such as light microscopy to assay structural features, cytological assays, immunofluorescent labeling of CD11B, CD18, CD64, CD66B, chemical change assays using flow cytometry, etc., transcriptomic signature analysis, etc.), suggesting an antibiotic regimen, suggesting an antiviral regimen, suggesting a chemotherapy regimen, informing a health care provider of the patient's health status, indicating the urgency of a health status, informing a patient of a health status, displaying a health status, and / or other appropriate interventions.

[0073] The method of the preferred embodiment and variations thereof may be embodied and / or performed at least in part as a machine configured to receive a computer readable medium storing computer readable instructions. The computer readable medium may be stored in any suitable computer readable medium, such as a RAM, a ROM, a flash memory, an EEPROM, an optical device (CD or DVD), a hard drive, a floppy drive, or any suitable device. The computer executable component is preferably a general or application specific processor, although any suitable dedicated hardware or hardware / firmware combination device may alternatively or additionally execute the instructions.

[0074] Embodiments of the systems and / or methods may include any combination and permutation of the various system components and various method processes, and one or more example methods and / or processes described herein may be performed asynchronously (e.g., serially), simultaneously (e.g., in parallel), or in any other suitable order by and / or using one or more example systems, elements, and / or entities described herein.

[0075] As those skilled in the art will recognize from the foregoing detailed description and drawings and the appended claims, modifications and variations can be made to the preferred embodiments of the invention without departing from the scope of the invention, which is defined in the following claims.

Claims

1. 1. A system for determining an immune activation state of a subject, the system comprising: - a microfluidic device comprising an inlet, an outlet, an alignment region configured to align particles in the microfluidic device, and a deformation region configured to deform particles; an image acquisition system, the field of view of which images the deformation region of the microfluidic device; and A processor, For each white blood cell from the subject's blood sample, determining one or more biophysical parameters of the white blood cells based on a plurality of images taken as the white blood cells traverse the deformation region; - adjusting one or more biophysical parameters based on at least one of the flow rate and the path traversed by the leukocytes; and determining an index for a set of white blood cells from a blood sample of the subject based on the adjusted one or more biophysical parameters; a processor configured to determine the immune activation state of the subject based on the index; and Including, the system.

2. The system of claim 1 , wherein if an image of white blood cells includes an overlapping white blood cell, the processor excludes the white blood cell from the set of white blood cells.

3. The system of claim 2 , wherein the processor is further configured to detect overlapping white blood cells based on monotonicity of characteristic motion between images of the white blood cells.

4. 2. The system of claim 1, wherein the processor is configured to determine, for each white blood cell, an entry position of the white blood cell into the deformation region and an exit position of the white blood cell from the deformation region, wherein if the entry position or the exit position differs from an expected entry position or an expected exit position, the white blood cell is excluded from the set of white blood cells.

5. 2. The system of claim 1, wherein the processor is further configured to determine, for each white blood cell, a cell type of the white blood cell, and wherein linearly adjusting the one or more biophysical parameters is further dependent on the cell type.

6. The system of claim 6 , wherein the processor is configured to determine cell type based on cell size and light intensity.

7. 6. The system of claim 5, wherein, when the cell type of the white blood cell is a neutrophil, linearly adjusting the one or more biophysical parameters further comprises adjusting the one or more biophysical parameters based on a z-plane of the white blood cell.

8. 2. The system of claim 1, wherein adjusting the one or more biophysical parameters based on the handedness of the pathway comprises linearly adjusting the one or more biophysical parameters when the pathway of the white blood cell includes a first handedness and not linearly adjusting the one or more biophysical parameters when the pathway of the white blood cell includes a second handedness.

9. The system of claim 1 , wherein the processor is configured to determine an index based on an average size of neutrophils and a 75th percentile of leukocyte trajectories within the deformation region.

10. The system of claim 1, wherein the immune activation state of the subject is not determined if the average velocity of the leukocytes through the microfluidic device is less than 1 m / s or more than 5 m / s.

11. 1. A method for determining the immune activation status of a subject, the method comprising: For each white blood cell from the subject's blood sample, - deforming said white blood cells within a deformation region of a microfluidic channel; - acquiring a plurality of images of the white blood cell as it traverses the deformation region; - determining one or more biophysical parameters of the white blood cells based on the plurality of images; - adjusting the one or more biophysical parameters based on at least one of a flow rate and a path traversed by the leukocytes; - determining an index for a set of white blood cells from a blood sample of the subject based on the adjusted one or more biophysical parameters; - determining the immune activation status of the subject based on the index; A method comprising:

12. The method of claim 11 , wherein if an image of a white blood cell contains an overlapping white blood cell, the white blood cell is excluded from the set of white blood cells.

13. The method of claim 12 , further comprising detecting the overlapping white blood cells based on monotonicity of characteristic motion between images of the white blood cells.

14. 12. The method of claim 11, further comprising the step of determining, for each white blood cell, an entry position of the white blood cell into the deformation region and an exit position of the white blood cell from the deformation region, and if the entry position or the exit position differs from an expected entry position or an expected exit position, the white blood cell is excluded from the set of white blood cells.

15. 12. The method of claim 11, further comprising, for each white blood cell, determining a cell type of said white blood cell, and wherein linearly adjusting said one or more biophysical parameters is further dependent on said cell type.

16. 16. The method of claim 15, wherein the set of white blood cells comprises a set of cells that is dependent on each cell type of the white blood cells.

17. 16. The method of claim 15, wherein, if the cell type of the white blood cell is a neutrophil, linearly adjusting the one or more biophysical parameters further comprises adjusting the one or more biophysical parameters based on a z-plane of the white blood cell.

18. 12. The method of claim 11, wherein adjusting the one or more biophysical parameters based on the flow path comprises linearly adjusting the one or more biophysical parameters when the path of the white blood cells includes a first handedness and not linearly adjusting the one or more biophysical parameters when the path of the white blood cells includes a second handedness.

19. The method of claim 11, wherein the index is determined based on the average size of neutrophils and the 75th percentile of the viscoelastic inertial response of monocytes.

20. 12. The method of claim 11, wherein the immune activation state of the subject is not determined if the average flow velocity of the leukocytes through the microfluidic channel is less than 1 m / s or more than 5 m / s.

Images