High speed microscopic imaging of objects moving along an optical axis

Abstract

Optical systems having a fluidic system with an incoming flow channel oriented along an optical axis, connected to an outgoing flow channel diverted away from the axis. A cross-sectional volume of the incoming channel can be illuminated so that a sensor focuses on objects within the illuminated volume, receiving light from the direction of the optical axis. Methods for using the optical system to detect or sort objects in the fluid flowing along the optical axis prior to diversion.

Description

High Speed Microscopic Imaging of Objects Moving Along an Optical AxisTECHNICAL FIELD

[0001] The present invention relates to the field of optics and high-throughput imaging of microscopic objects.BACKGROUND OF THE INVENTION

[0002] Conventional setups for imaging moving microscopic objects of interest — such as those used in flow cytometry — are oriented such that the viewing direction is perpendicular to the object’s primary direction of motion (in other words, medial to the flow direction). The effect is that objects pass from one end to another, laterally across a camera’s two-dimensional frame or field of view. While intuitive to set up in theory (e.g, by positioning an objective lens perpendicular to a capillary tube or to a length of a microfluidic channel), in practice setups require precise positioning of objects in fluidic channels relative to the focal plane of the optical system, whose depth of focus is much smaller than the channel dimensions (e.g, 1-20 pm for high numerical aperture [NA] objectives). To mitigate the alignment challenge, flow cytometers often employ fluidic operational mechanisms such as hydrodynamic focusing within narrow channels (via sheath flow, inertial focusing, channel restrictions, etc.). Additionally, fabrication of the channels and datuming structures on the disposables can require tight dimensional specifications. The components within the optical system can also be strictly aligned, further requiring high-toleranced structures andmanually intensive procedures. Altogether, current approaches present several drawbacks, including:

[0003] Cost. Due to the complexities and tolerances required to achieve proper alignment, the costs of consumable fabrication, fine- quality hardware, and assembly processes are relatively high.

[0004] Data richness. Most systems only provide basic counts of detection events. More detailed characterizations (e.g, morphological / morphometric characterizations, size, etc.) from opticalbased flow setups are often unachievable, requiring prohibitively longer dwell times (z.e., lower flow rates or longer camera exposure times) and significantly lowering the throughput.

[0005] Throughput. Systems geared toward binary detection events operate at relatively high flow rates compared to other microfluidic applications (e.g, 100-1000 pL / min), but still typically require sequential, one-by-one observations. The detection events are difficult to parallelize and scale up, owing to hardware complexities surrounding the detection zone.

[0006] Dynamic range. The restrictive depth of focus (z.e., spatial range and resolution) and exposure times (z.e., temporal resolution) of current setups imposes limitations to the range of object sizes, concentrations, and flow rates able to be accommodated.

[0007] Reliability. The narrow flow channels required of existing setups often lend themselves to partial or complete channel blockages (z.e., run failures) caused by intermittent but inevitable dust or debris present in input samples. Small channel geometries are also more prone to fabrication, tolerancing, and quality control issues.SUMMARY OF INVENTION

[0008] The present invention provides a method for rapidly imaging a population of microscopic objects dispersed in a fluid medium. The invention uses an optical imaging detector whose plane of view focuses on the cross section of an orifice through which the objects flow. The invention provides an orientation of the optical path that faces the incoming or outgoing primary flow path of the objects, along its axial direction. A representative imaging system may be constructed to have a sufficient acquisition rate, field of view, data processing speed, and memory to detect and assay all targets within a set period of time and sample flow rate.

[0009] The invention may be applied to target detection, enumeration, sorting, screening, or other multi-parametric analyses, including fluorescence-activated expression, morphological characterization, and other applications typical of microscopic imaging systems. The invention may generally have utility in the life sciences (involving biological cells, viruses, vesicles, multicellular organoids / spheroids, pollen, etc.), chemistry and materials sciences (involving emulsions, colloids, micro- and nanoparticles, etc.), or fluid mechanical studies (involving particle imaging velocimetry, etc.). It can have broader impacts in biomedical applications (oncology, hematology, serology, therapeutics, tissue engineering, etc.), environmental monitoring (aerosols, pollutants, water contamination, etc.), and as a tool for manufacturing processes (quality control or real-time characterization of particle synthesis, crystallization, etc.), among others.BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 shows a simplified schematic illustrating representative elements of the invention with their respective orientation. In this embodiment, an incoming flow (1) carries a suspension of objects of interest (2) in a fluid medium (12), which flows through an optically transparent microfluidic flow channel (3). An optical system is placed external and near to the microfluidic channel, as follows. The optical system is oriented such that an objective lens (6) is aligned to the primary optical and flow axes (11) of the flow channel, and is focused within the flow channel’s cross-section, and views the incoming (or “oncoming”) flow of objects through its cross-section, moving toward the sensors. The channel is T-shaped to divert one or more outgoing flows (5a, 5b) away from the objective lens, allowing clearance for positioning the optical system. A planar light sheet (4) illuminates the channel and objects at the focal plane of the objective lens. The scattered light and emission spectra from the illuminated objects are collected by the objective, passed through an optical emission filter (8) and a transmitted tube lens (9), and are captured by a fluorescence event camera sensor (10). In addition, a dichroic beamsplitter (7) is situated between the objective lens and the emission filter to reflect selected wavelengths along a reflected optical axis (Ila), having an aligned reflected tube lens (9a) and a scattering camera (14).

[0011] Figure 2 shows an expanded optical system that is configured with multiple cameras and optical filters for selective detection of multiple fluorescence emission spectra. As in Figure 1, incoming oroncoming flow (1) carries a suspension of objects of interest (2) in a fluid medium (12) through a flow channel (3).

[0012] An objective lens (6) is aligned to the primary optical and flow axes (11) of the flow channel, and the lens is focused within the flow channel’s cross-section, where the focal plane can be illuminated by a planar light sheet (4). The objects of interest flow through the crosssection of the flow channel, moving toward the sensors and is diverted through a T-shaped channel in one or more outgoing flows (5a, 5b) away from the objective lens. The scattered light and emission spectra from the illuminated objects are collected by the objective. A long pass dichroic beamsplitter (7a), selectively reflects certain (relatively shorter) wavelengths in the direction of a first reflected optical axis (Ila) through a first reflected tube lens (9a) toward a scattering camera (14a).

[0013] The wavelengths not reflected by the long pass dichroic beamsplitter are then selectively reflected through a relatively longer wavelength dichroic filter (7b) in the direction of a second reflected optical axis (11b) through a second reflected tube lens (9b), a first emission filter (8b) and a first fluorescence camera (14b).

[0014] The remaining wavelengths are then selectively reflected through a still longer wavelength dichroic filter (7c) in the direction of a third reflected optical axis (11c) through a third reflected tube lens (9c), a second emission filter (8c) and a second fluorescence camera (14c). Additional sets of dichroic filters, tube lenses, emission filters, and cameras can be added for desired, typically increasing, wavelengths (13).

[0015] Finally, remaining wavelengths are reflected by a mirror (15) in the direction of a wth reflected optical axis (lln) through an / 7 threflected tube lens (9n), an (w-l)th emission filter (8n) and an (w-l)th fluorescence camera (14n). This configuration allows for the interrogation and profiling of multiple dyes or other fluorescence characteristics from the objects of interest.DETAILED DESCRIPTION OF THE INVENTION

[0016] The invention is an imaging system whose primary optical axis lies along the primary direction of motion of the object(s) of interest. In other words, the detector’s field of view faces the targets’ oncoming or outgoing direction of motion.

[0017] The invention provides an optical system having an optical axis, having a fluidic system with an incoming flow channel (3) for incoming flow (1) oriented along the axis, an outgoing flow channel oriented away from the axis, and a fluidic connection between the incoming channel and the outgoing channel. The system can have a light source that projects a planar light sheet (4) to illuminate a cross-sectional volume of the incoming channel. The system can also include a Laboratory Information Management (LIMS) system to manage the samples and associated data. Important aspects to the invention are as follows:Sample Handling— Sample

[0018] The methods of the invention can be used to assay a sample. As used herein, a “sample” can be a mixture of objects of interest (2) dispersed in a fluid medium. The fluid medium (12) may consist of onemore liquids and / or gases. Biological samples can be derived from viruses, bacteria, microorganisms, plants, and animals, such as humans. The sample can be of a body fluid (e.g, blood, cerebrospinal fluid, peritoneal fluid, pleural effusion, nasal swab, saliva, etc.), cells, or cell- free fractions isolated from a sample, or a tissue sample. The sample can also be from a culture medium. The sample may be dispersed in or among one or more liquids and / or gases, such as a buffered aqueous solution, oil, or air. The sample mixture can be of a solution, a colloid, or a suspension.

[0019] The sample can be a homogeneous mixture or a heterogeneous mixture, whereby a portion of the objects in the sample are the target objects of interest. Prior to use with the invention, the sample may be pre-treated or specifically labeled to amplify signal during imaging (e.g, via immunofluorescence, quantum dots, or chemical or enzymatic colorimetric staining, etc.). The target objects of interest can be microscopic. As used herein, “microscopic” refers to objects that are resolvable after magnification (e.g, exactly or about lx, 1.5x, 2x, 2.5x, 3x, 4x, 5x, 6x, 7x, 8x, lOx, 15x, 20x, 25x, 30x, 40x, 50x, 60x, 70x, lOOx, 150x, 200x, 250x, 300x, 400x, 500x, 750x, or lOOOx or more, or in a range between any of the preceding numbers).— Orifice

[0020] As used herein, an “orifice” is a region that confines the sample within a cross-sectional area (e.g, exactly or about 10'18m2, IO’17m2, 10'16m2, 10'15m2, 10'14m2, 10'13m2, 10'12m2, 10'11m2, IQ-10m2, 1 Q-9m2, 10-8m2, 10-7m2, 10-6m2,Qr10-5m2,Qr more,orjn arange between any of the preceding numbers) and through which the sample may pass. The sample is imaged within an orifice near to the sizeof the field of view, such that all objects of interest passing through may be detected. The optical path to the orifice is optically transparent. The shape of the orifice can be rectangular, rounded, elliptical, triangular, or other shape. The orifice can be an opening or a cross-sectional region of a greater flow channel and fluidic system. As used herein, a “flow channel” is a confined pathway through which a fluid can be transported. A flow channel can be derived from a glass capillary tube, extruded plastic tubing, or an injection-molded thermoplastic microfluidic device. A greater fluidic system may contain one or more incoming (1) or outgoing (5a, 5b) flow channels, chambers, or other features to contain, interrogate, or route the sample. After optical interrogation, samples may be diverted away from the imaging axis (z.e., to provide sufficient clearance for the objective lens). This may be achieved using common flow channel geometries such as a T-shape (1, 5a, 5b), L-shape, or Y-shape, or by three-dimensional counterparts (e.g , cone-shaped). Flow channels may also utilize sharp angles or curved surfaces.— Motion driver

[0021] The system can have a motion driver that is configured to generate fluidic flow within the flow channel. The sample can pass through the imaging plane, and the motion may be generated by various means, including flow pumps (e.g, rotary, positive-displacement, peristaltic, diaphragm, etc.), pressure-driven pumps (e.g, compressed air, vacuum, etc.), passive pumps (e.g , capillary action, siphoning, etc.), or other means. The average velocity of the objects of interest passing through the imaging plane can be on the order of, exactly, or about1 O'9m / s, 1 O'8m / s, 1 O'7m / s, 1 O'6m / s, 1 O'5m / s, 1 O'4m / s, 1 O'3m / s, 1 O'2m / s, I O’1m / s, 2.5 / I 0’1m / s, 5 / I 0’1m / s, 7.5 / I 0’1m / s, 1 m / s, lO m / s, 102m / s, ormore, or in a range between any of the preceding numbers. The motion driver performance and response time may be augmented with auxiliary hardware and feedback, such as pressure regulators or PID (proportional-integral-derivative) controllers. Additional microchannel geometries or external hardware may aid in generating unique flow profiles and behaviors (e.g, ultrasonic-induced vortices and recirculating flow, pulsed or oscillatory flow, or extensional / compressional flow profiles that induce object deformation, etc.). Additional, independent motion drivers may be added downstream from the sensing region to implement a sorting mechanism (e.g, via dielectrophoresis (DEP), acoustofluidics, or pneumatic valving, etc.). Alternatively, the sample can be stationary while a relative motion is achieved from the imaging setup (e.g, an objective lens moving on a stage). Finally, the speed of motion or flow rate can be adjusted to maximize throughput while maintaining capability to optically detect the objects.Imaging Setup— Sensor

[0022] The optical system can further have one or more sensors that are positioned to receive light from the optical axis, where the sensor can focus on the illuminated volume. Different sensors can be employed for simultaneous detection of dependent or independent characteristics (e.g, fluorescence, brightfield, etc.). Examples of sensors include a scattering camera (14, 14a) and fluorescence cameras (14a, 14b, 14c, 14n).

[0023] Important to the operation is the selection of a sensor to detect individual objects passing through the imaging plane (events). The sensor pixel size and / or pitch can be selected, in combination with thelevel of magnification, to detect microscopic objects of interest. The overall sensor size, in combination with the level of magnification, can be selected to encompass the size of the orifice or a subsection of the orifice. The resulting field of view may resolve single or multiple detection events simultaneously.

[0024] The sensor orientation defines the optical axis. As used herein, “optical axis” is the primary intended path (11) along which light from the objects of interest travel to the sensor, whether through or from a series of other optical elements. Unlike traditional imaging setups in which the detection time window is primarily limited by the field-of- view size as objects pass from one end to the other, the analogous time window of the invention is the relatively shallow depth of focus. Since the depth of focus in microscopic applications is typically smaller than a field-of-view dimension, the sensor can have a relatively higher frame rate (z.e., temporal resolution) to detect a moving object passing through the detection zone in a given period of time (e.g, on the order of, exactly, or about 10, 102, 103, 5*103, 104, 2*104, 5*104, 105, 2*105, 5><105, or 106or more equivalent frames per second [FPS], or in a range between any of the preceding numbers). The sensor can be a high- or ultra-high speed camera, which are typically built around CMOS (Complementary Metal Oxide Semiconductor), such as an sCMOS (scientific Complementary Metal Oxide Semiconductor) and / or CCD (charge-coupled device) sensors. The sensor can be an event-based sensor (10), also known as a temporal contrast sensor, neuromorphic camera, silicon retina, or a dynamic vision sensor, which has characteristically high sampling rates (tens-, hundreds-, thousands-, tens- of-thousands- to hundreds-of-thousands equivalent FPS), which enable orders of magnitude greater frame rates than those of traditional cameras(up to the low hundreds FPS). With progressing advancements in sensor sizes i.e., field of view), speeds (e.g, related to bandwidth, frame rate, readout, transfer, etc.), and availability, high-speed cameras provide sufficient temporal resolution to offset narrow depths of focus and overcome fundamental limitations of high-speed axial detection.— Illumination

[0025] Objects can be sufficiently illuminated at the focal plane of the imager to provide enough signal for detection. While traditional bright- field or reflected light illumination schemes may suffice, a light sheet, also known as Selective Plane Illumination Microscopy (SPIM) can be utilized to minimize background and signal from out-of-plane objects, while illuminating the entire field-of-view and minimizing photobleaching. Thus, a light source can be provided that is capable of projecting a light sheet. As used herein, a “light sheet” is an imaging method in which a thin, planar light source illuminates a sample perpendicularly to the direction of observation. The light source can be an arc lamp, a halogen bulb, a laser diode, or a light emitting diode.

[0026] The light source can be structured by multiple means to improve image quality and / or object detection and recognition. Light entering or leaving the imaging plane can be spatially oriented and / or patterned (e.g, via Bessel beams, Airy beams, speckle patterns, optical lattices, spatial-frequency banding, two- or multi-photon excitation, etc.), whether to improve uniformity, support deconvolution and superresolution imaging, or to incorporate image encoding and / or decoding information. Light intensity can be temporally modulated to increase signal-to-noise i.e., increasing light intensity) while minimizing photodamaging effects. In one example, the illumination can be strobed(i.e., blinking, or turned on and off) and / or swept / scanned (z.e., also increasing illumination area) at high frequencies (e.g., durations within a range between femtoseconds to nanoseconds to microseconds or more), which increases the number of detection events of event camera sensors. The light source can also provide light that is polarized, where the polarization can be linear, circular, or elliptical. In some embodiments, the light sensor can be configured to detect changes in the polarization of the light. Such changes can be due to the properties of the objects of interest in the sample.— Magnification

[0027] One or more lenses can be employed to resolve the target microscopic objects. Setups may utilize long working distance (LWD) objective lenses for increased flexibility in containing or routing the samples around the optical hardware. Objective lenses (6) used in the setup can have a numerical aperture (NA) exactly or about 0.25, 0.3, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85 or more, or in a range between any of the preceding numbers. During operation, objective lenses can be dipped or immersed in air {i.e., dry), in water, or in oil. Objective lenses can be selected for phase contrast (Ph) imaging, bright- field (BF) imaging, dark-field (DF) imaging, fluorescence (FL) imaging, differential interference contrast (DIC), or other microscopic imaging techniques.— Other optical elements

[0028] A tube (or converging) lens may be positioned on the primary axis (9) or on the reflected optical axes (9a, 9b, 9c, 9n) to form images for the various cameras (10, 14, 14a, 14b, 14c, 14n).

[0029] Various optical filters (8, 8b, 8c, 8n) (e.g., dichroic, bandpass, neutral density, acousto-optic tunable filters [AOTFs], spatial light modulators [SLMs], etc.) and / or beamsplitters (7, 7a, 7b, 7c) may be utilized for both the illumination side and the detection side to modulate (z.e., intensity / amplitude, wavelength / frequency, phase / coherence), direct, and / or target one or more specific types of signal e.g., sidescattering, fluorescence detection). Fluorescence detection may target light wavelengths below or beyond the range of ultraviolet (UV), visible, or near or far infrared (IR) spectrums. Motion stages may also be used for run-to-run and / or dynamic focus repositioning or 3D profiling of a sample container. The optical system can include one or more beamsplitters, for example in line with the flow (primary) optical axis (11) so that it is capable of redirecting a portion of light along a reflected (or semi-reflected) optical axis (Ila) in another direction. As shown in Figure 2, the system can have combinations of multiple beamsplitters (7a, 7b, 7c) and / or mirrors (15) that direct light along multiple corresponding optical axes (Ila, 11b, 11c, lln). In other configurations, a path of redirected light can itself be further redirected by one or more dichroic beamsplitters leading to further optical paths. For example, one beamsplitter can split light to two other beamsplitters, leading to four paths / cameras in an essentially branched configuration.Cocktails and Kits

[0030] The invention also provides cocktails of a carrier fluid that is compatible with the fluidic system. The cocktail can also have fluorescent labels capable of excitation by the light source.

[0031] The invention further provides kits having the cocktail and a washing fluid, for example. The kit can also have a microfluidic chip with a reservoir that is preloaded with the cocktail.Methods

[0032] The invention also provides methods for detecting objects in a flowing fluid along the optical axis. The methods can include the steps of providing a flow of fluid medium containing objects of interest into the incoming channel of the optical system; projecting a light sheet that illuminates a cross-sectional volume of the incoming channel along the optical axis; and / or capturing images of objects that are illuminated by the light sheet.

[0033] The method can also include diverting the fluid medium away from the optical axis through the outgoing flow channel. The method can further include sorting objects in the fluid medium according to one or more predetermined criteria, for example by fluorescence intensity or wavelength, or by size, shape, or other physical property. The selected objects and immediately surrounding fluid medium can then be redirected to a different flow channel for segregation, collection, or further analysis or selection, for example.Algorithms

[0034] The invention provides computer-performable algorithms for performing the analysis of the steps of the method. Some of the steps of the algorithm can be stored or performed in memory and / or a hardware accelerator (e.g, field-programmable gate array [FPGA], graphicsprocessing unit [GPU], neural processing unit [NPU], frame grabber, etc.), or a processing unit from cloud-based memory, a mobile electronic device, and / or a portable memory storage unit. The output can be transmitted to a mobile device.

[0035] Image processing algorithms are developed to effectively detect objects from raw signals, unique to the optical setup. Algorithms may take into account multidimensional data streams from multiple event cameras, effectively interpreting the appearance and disappearance of signal into and out of the illumination and field of view. Signal processing (e.g, object segmentation, deblurring, frame differencing, trajectory tracking, morphometric classification, etc.) is customized to object detection, localization, discretization, and characterization. Object recognition pipelines can also be augmented by convolutional neural networks (CNNs), support vector machines (SVNs), and / or other models within machine learning (ML). Resulting output metrics can include quantification of object quantity, speed, position, trajectory, fluorescence emission / expression, morphology, and other metrics for each detection event. The resulting output metrics can be linked to and trigger additional software-, firmware-, and hardware-driven steps, such as fluidic perturbations near or downstream from the detection site to redirect specific objects to a different flow channel.Advantages— Practical use

[0036] Ease of focusing: Since the field of view may be tuned to cover the entire cross-section of a channel or orifice, the setup naturally ensures objects will pass through the focal plane and field of view atsome point in time. This reduces or eliminates the need for operational hydrodynamic or objective focusing and significantly relaxes alignment and tolerancing requirements within a microfluidic channel.

[0037] Flexibility: Once set up, imagers are readily adjustable simply by tuning the field-of-view and / or orifice size without needing significant re-optimization.— Performance improvements

[0038] Sensitivity: In theory, all objects of interest may be realistically detected so long as the temporal resolution of the event camera reasonably accommodates the target objects’ size and speed. This enables the detection of ultra-rare events. Combined with sufficient signal and data processing speed that trigger downstream flow perturbations — achievable using common fluidic hardware — detected objects of interest may be redirected and sorted from the bulk sample, achieving high capture efficiency, purity, and enrichment.

[0039] Cost: The simplification of focusing enables significant cost reduction of both the hardware and consumables.

[0040] Data richness: Owing to the high acquisition rates of event cameras, significantly more data may be generated from the spatiotemporal data streams (e.g, morphology, spacing, positioning, sizing, etc.) compared to traditional binary event detection. Combined with algorithmic reconstruction, three-dimensional morphologies may even be extracted for each object of interest.

[0041] Throughput: The invention is indifferent to having multiple objects appear in one field of view, thus enabling higher concentrations of sample to be processed, significantly improving throughput and efficiency.

[0042] Dynamic range: The field of view and high camera speed combined can accommodate a greater range of detectable object sizes (e.g, vesicles to single cells to organoids), speeds, and concentrations.

[0043] Reliability: The method’s imaging scheme enables orifice or channel characteristic cross-sectional length scales greater than 20, 50, 100, 200, 500, or 1000 micrometers, which is wider than typical contaminating dust and debris, thus minimizing run failures due to channel blockages.— Broader impacts

[0044] Life science research: Examples include the analysis of heterogeneous populations (including tumor spheroids, organoids, or stem cells), fluorescence-activated cell sorting (FACS), precision singlecell dispensing, and marine biology research.

[0045] Biomedical: Examples include rare cell enumeration of liquid biopsies (e.g, circulating tumor cells, exosomes), deformability cytometry for the detection of sepsis (e.g, functional states of white blood cells), quality control of extracellular vesicles for therapeutics, nucleic acid amplification testing (NAAT; e.g, droplets in digital polymerase chain reaction [PCR]), pharmacokinetic and pharmacodynamic responses on the immune system including cytotoxicity, and hemocytometry for quantifying a complete blood count (CBC).

[0046] Materials: Examples include nanoparticle synthesis, sizing and sorting, optimization of crystal growth and formation, or other industrial production applications (e.g, high-speed monitoring of inkjet particles, fuel atomization, or detection of wear particles in lubricants).5

[0047] Environmental: Examples include the analysis of bioaerosols, wastewater contamination, or air pollution.ExamplesExample 1: Diagnosis of Septic Infection

[0048] Sepsis of the blood is a potentially fatal infection and typically requires a time-consuming blood culture to gauge severity. A device that rapidly diagnoses sepsis benefits such environments, where patients at risk of septic shock should be rapidly identified and treated. The invention is applied two-fold, whereby (1) a high-magnification optical setup directly visualizes and enumerates the presence of bacteria in a blood sample, and potentially identifies an initial classification based on morphology (cocci, bacilli, or spirochaetes); or (2) in combination with a microfluidic channel generating an extensional flow profile, the sensor rapidly detects signatures in the host immune response (for example, by the measurement of deformability or other biophysical characteristics activated immune cells within the blood) as a secondary indicator for sepsis.Example 2: Liquid Biopsy Device for Determining Cancer Patient Prognosis

[0049] In metastatic cancer, cells from a primary tumor shed into the bloodstream and are capable of spreading to a secondary site and worsening patient outcomes. The number of circulating tumor cells (CTCs) in the bloodstream correlates with patient prognosis and can be used as a metric for patient response to treatment, such as from chemotherapy. However, CTCs are exceedingly rare, as few as one per milliliter of blood, and are thus challenging to detect among red and white blood cells numbers that can be orders of magnitude greater.Current solutions rely on time-consuming enrichment processes by sizebased filtration methods or antibody capture, or by imaging across expansive glass slides. Using immunofluorescent labeling of blood samples that target epithelial biomarkers characteristic of CTCs or other specific phenotypes, the invention rapidly screens through high volumes of blood and efficiently detects and enumerates CTCs with high sensitivity. The system also distinguishes individual CTCs from CTC clusters, and is extended to detect extracellular vesicles or exosomes tagged with tumor-specific markers, offering broader liquid biopsy capabilities.Example 3: Purification Device Implementing Fluorescence- Activated Cell Sorting (FACS)

[0050] The invention is fitted in-line with a sorting fluidics architecture, such as those used by fluorescence-activated cell sorting to rapidly quantify and purify subpopulations of labeled cells in a mixture. In addition to the traditionally offered quantifications — which include a simple count and fluorescence intensity measurement (z.e., phenotypic expression levels) — the invention provides significant additional morphological information, including size, shape, cell-to-cell binding, etc. As outlined previously, the invention further increases throughput, dynamic range, and sensitivity of existing instrumentations, which is especially crucial for rare subpopulations and rapid single-cell analysis applications. In certain embodiments, label-free classification is achieved using high-resolution morphological features such as nuclear- to-cytoplasm ratio, granularity, or deformability metrics, enabling sorting of fragile or poorly characterized populations.Example 4: Air Quality Measurement Device

[0051] The invention is applied for the detection and identification of airborne microparticles in the field of air quality monitoring to assess risk and potential harm in both outdoor and indoor environments. The imaging technique is tuned to recognize morphological features and to classify targets such as soot or other particulate matter generated by pollutant-generators (e.g, cigarette, vehicle, and industrial emissions) or natural disasters (e.g, forest fires), and allergens (e.g, pollens, dander, and mold spores). Optical signatures (e.g., fluorescent or scatter) also enable efficient detection of airborne droplets containing sub-micron nanoparticle or volatile organic compound (VOC) targets. Additionally, the invention is applied to quantify and purify micro-scale liquid bioaerosol droplets — potentially harboring precursors of respiratory infectious diseases (z.e., viral, bacterial, or fungal pathogens) — particularly from those emitted by individuals in crowded indoor environments via coughing, sneezing, talking, high-intensity activity, etc. In mobile or aerial implementations, such as drone-mounted systems, the invention may be used for wide-area sampling and real-time geospatial mapping of pollutant and allergen distributions.

[0052] The headings provided above are intended only to facilitate navigation within the document and should not be used to characterize the meaning of one portion of text compared to another. Skilled artisans will appreciate that additional embodiments are within the scope of the invention. The invention is defined only by the following claims; limitations from the specification or its examples should not be imported into the claims.

Claims

1. CLAIMSI claim:

1. An optical system having an optical axis, comprising a fluidic system comprising an incoming flow channel oriented along the axis, an outgoing flow channel oriented away from the axis, and a fluidic connection between the incoming channel and the outgoing channel; and a light source capable of illuminating a cross-sectional volume of the incoming channel.

2. The apparatus of claim 1, wherein the fluidic system further comprises a second outgoing flow channel.

3. The apparatus of claim 2, wherein the incoming flow channel is connected to the outgoing flow channels by a T-junction.

4. The apparatus of claim 2, wherein the incoming flow channel is connected to the outgoing flow channels by a Y-junction or a cone junction.

5. The apparatus of claim 1, wherein the light source is capable of projecting a light sheet.

6. The apparatus of claim 1, wherein the optical system further comprises a sensor that is positioned to receive light from the optical axis, and wherein the sensor is capable of focusing on the illuminated volume.

7. The apparatus of claim 6, wherein the sensor is selected from the group comprising an event-based sensor, an sCMOS, and a CCD.

8. The apparatus of claim 1, wherein the optical system further comprises a component selected from the group consisting of a wavelength filter, a dichroic mirror, a prism, a notch filter, and a bandpass filter.

9. The apparatus of claim 1, further comprising a beamsplitter in line with the optical axis that is capable of redirecting a portion of light in another direction.

10. The apparatus of claim 1, further comprising another beamsplitter and another sensor.

11. The apparatus of claim 1, further comprising a motion driver that is configured to generate fluidic flow within the flow channel.

12. The apparatus of claim 1, further comprising a LIMS system.

13. A cocktail comprising a carrier fluid that is compatible with the fluidic system and fluorescent labels capable of excitation by the light source of claim 1.

14. A kit comprising the cocktail of claim 13 and a washing fluid.

15. A kit comprising a microfluidic chip with a reservoir that is preloaded with the cocktail of claim 13.

16. A method for detecting objects in a flowing fluid along an optical axis, comprising the steps of(a) providing a flow of fluid medium containing objects of interest into the incoming channel of the optical system of claim 1;(b) projecting a light sheet that illuminates a cross-sectional volume of the incoming channel; and(c) capturing images of objects that are illuminated by the light sheet.

17. The method of claim 16, further comprising(d) diverting the fluid medium away from the optical axis through the outgoing flow channel.

18. The method of claim 16, further comprising(e) sorting objects in the fluid medium according to predetermined criteria.

19. A computer-performable algorithm for performing the analysis of step (c) of the method of claim 16.

20. The algorithm of claim 19, wherein at least one of the steps is stored or performed in memory or a processing unit from the group consisting of cloud-based memory, a mobile electronic device, and a portable memory storage unit.

21. The algorithm of claim 19, wherein output is transmitted to a mobile device.

Images