Contaminant sensing interrogator (CSI) for biologics and complex fluids

Abstract

Biopharmaceutical products require sterility testing to ensure their safe use as injected medicines. The present invention performs this task automatically. By testing the entire fluid volume using optical sensors with no blind spots, the system described herein counts and measures both microbial and subvisible particle contaminants immediately and without modification of the tested products. Actionable data collected by the system includes classification or identification of contaminants and other objects via light-based measurements at the level of each single cell, colony forming unit, or particle.

Description

[0001] Contaminant Sensing Interrogator (CSI) for Biologies and Complex Fluids

[0002] FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT

[0003] This invention was made with government support under Contract No. HR001122C0147 awarded by the Defense Advanced Research Projects Agency, Small Business Programs Office (SBPO). The government has certain rights in the invention.

[0004] CROSS-REFERENCE TO RELATED APPLICATIONS

[0005]

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 511,240 entitled, "METHOD AND APPARATUS FOR RAPID STERILITY TESTING OF BIOLOGICAL PRODUCTS," filed June 30, 2023, which is hereby incorporated by reference in its entirety.

[0006] BACKGROUND

[0007]

[0002] Discerning whether sparse contaminant entities reside within complex fluids is a challenge. Standard methods of filtering, plating, and culturing to check for sterility are time consuming processes not readily adaptable to automation. Current sterility testing is expensive in terms of both time and financial resources, requires skilled personnel and specialized facilities, and is not amenable to in-line or continuous processes. If these standard methods could be tested against and replaced by an automatic fluid testing strategy, complex fluids, including valuable biopharmaceutical or biomanufactured products, could be rapidly inspected immediately and in situ for quality control using small fluid volumes and without human labor or exogenous agents.

[0008]

[0003] We herein disclose a solution that provides significant improvements compared to current sterility testing methods, addressing these issues with a compact modular instrument consisting of multiple, information rich sensors for contaminant detection and classification. CSI (the Contaminant Sensing Interrogator) is an automated microscopy and microfluidics platform for rapidly examining sterility of complex fluids including bioreactor products, and can generalize to a wider range of use cases including water quality testing or testing of biofluids.

[0009]

[0004] CSI offers a strategy to address the additional challenge of discerning whether a putative contaminant organism is viable; only viable entities represent colony forming units (CFUs). Our sterility testing device based on light sensing detectors provides several data cross sections of a detected object. It then utilizes these multiple components of information to construct a more complete picture of the object and classify the object type or identity. This is done either by a direct, rational approach using automated physical analysis of the output data components; or alternatively, an artificial intelligence system uses the multicomponent output of the device as training data for machine learning, potentially using information in ways that would not be possible by a direct approach. In one example, the set of spectrographic records of a known contaminant particle in a variety of fluid environments could be used by physics guided Al as a reference data type for finding the same type of particle by matching spectrographs in additional fluid environments or samples. This Al capability offers the further advantage of a detection system that responsively adapts to different circumstances or improves over time.

[0010]

[0005] CSI is built to detect CFUs at simultaneously high specificity and sensitivity, perform these measurements automatically using small sample volumes, and within a short measurement duration. Detection of contaminant particles or CFUs is based on darkfield imaging of fluids steered through a microchannel. Due to the different size regimes of product components vs contaminant organisms or subvisible particles, scattered light can sensitively detect these organisms or particles within various complex fluid mixtures.

[0011]

[0006] The system views an area that includes the entire channel width (0.2-5 mm), and the fluid is imaged through the full channel depth (0.05-0.2 mm). As complex fluid products or biologies flow through the channel, any putative CFU residing in the fluid must pass three checkpoints along this obligate path. These include:

[0012] • Checkpoint 1 - Inlet Trigger

[0013] • Checkpoint 2 - Motion Analyzer

[0014] • Checkpoint 3 - Spectrographic Classifier

[0015]

[0007] In a compact apparatus, the three checkpoints of CSI work together to enhance the sensitivity and specificity of contaminant detection. Further, the information gathered at these checkpoints encompasses several data cross sections to discriminate CFUs from other non-CFU objects or particulates, or to identify or classify CFU types based on training of the system by machine learning.

[0008] Flow rate control permits object steering along the fluid path, and the ability to rapidly stop the fluid, reverse direction, and position a putative CFU for further examination under stopped flow conditions. Controlled fluid flow thus serves as a "conveyor belt" to move objects through the three machine-vision checkpoints of CSI. CSI Machine Vision Checkpoints

[0016]

[0009] At Checkpoint 1, the fluid is flowing at a constant rate (0.1-10 il / s) as it carries CFUs through the channel and each CFU entering Checkpoint 1 is seen as a streak. Streak length per camera exposure time gives a measurement of the CFU speed, which depends on its depth in the channel - a feature of Poiseuille flow.

[0017]

[0010] For streaks shorter than the light sensing region (0.2-5 mm in the flow direction), the full streak is in view and the total light in each streak is independent of streak length or object speed. Checkpoint 1 software adds the light along each streak to trigger the system to stop fluid flow when a CFU enters the viewing area.

[0018] [Oil] At Checkpoint 2, the fluid is stopped so that objects can be tracked to ascertain details of CFU diffusion or swimming. In the purely diffusive case, the Checkpoint 2 motion analyzer returns estimates of diffusion that are related to object size and viscosity of the surrounding fluid. In the case of active swimming of CFUs (bacteria), the motion analyzer tracks CFU trajectories to determine active motion parameters such as excursion length, swimming speed, pausing, and direction changes.

[0019]

[0012] At Checkpoint 3, spectrographs of the CFU are measured and separated from light interactions with the surrounding fluid. Spectrographs may be unique to each CFU species or type. We have successfully tested our system to measure single-CFU spectrographs using broadband scattered light and fluorescence at Checkpoint 3. Measurements based on absorbance or reflectance are among the additional modes of operation that could be used.

[0020] CSI Illustrated in a First Specific Device

[0021]

[0013] The first CSI hardware instance herein described comprises a compact set of core components that would easily fit in a lunchbox. This version of the device, designed for use in ambulatory or remote bioreactor facilities, is rugged with solid construction and no wear parts.

[0022]

[0014] In one embodiment, an apparatus utilizes a microfluidic flow channel with the entire volume of fluid passing through the channel depth and width visible by a multitude of widefield optical sensors for monitoring and detection, including at least one of a photon counting camera or neuromorphic event camera featuring low latency, a darkfield or scattered light sensor featuring sparse data and a high signal to background, and an imaging spectrometer featuring wavelength dependent intensity data capture. Each sensor in this system can count subvisible particulates including cells or microbial contaminants during rapid fluid flow. The operation of multiple sensors can be serial or simultaneous to enhance fidelity of detection via redundancy and cross checks. None of the optical components in this instance require changing focus or field of view to measure the entire fluid volume passing through the channel.

[0015] The flow channel is wider than it is deep. The shallow depth facilitates uniform detection and measurement of objects throughout the depth of fluid, whereas the wide viewing area across the channel facilitates high volume flow rates at a reduced linear speed of objects carried by the fluid in the flow direction. This design optimizes measurement sensitivity and throughput.

[0023]

[0016] Under nominal conditions the solution being evaluated will be free of contaminants and the testing system can operate under continuous uninterrupted flow. Each of the triple redundant sensors can detect a subvisible object under rapid fluid flow, triggering the system to make an immediate decision whether to count the object as a contaminant. If ambiguity exists and closer inspection is needed, the system can automatically stop the fluid flow and use fluid steering to hold the object for further inspection.

[0024]

[0017] We provide the following descriptions (A-E) as specific examples of inspection modes that can be used in this instance or in other instances of CSI hardware, noting that other methods can additionally or alternatively be used, and these examples are illustrative rather than limiting.

[0025] A. Optical Scatter Measurements Utilizing Light Sources that Illuminate the Channel from Various Directions

[0026]

[0018] Design implementations for the system can employ one or multiple light sources strategically positioned to illuminate the microfluidic channel from various directions, thereby enabling a comprehensive optical characterization of objects or particles within the channel. For instance, an overhead top-down illumination allows for the characterization of the optical properties of the object in a "forward" direction, capturing how light scatters off the surface when it is illuminated from above. Conversely, a bottom-up illumination from beneath the channel provides a "reverse" direction characterization, revealing how light interacts with the object when illuminated from below. This bidirectional illumination setup along with the optical collection of the corresponding microscope objective(s) and components enables the system to measure portions of and / or the entire total integrated scatter (TIS) from the object under various illumination conditions, significantly enhancing the detection and classification accuracy of particles. Furthermore, the system can measure and characterize the bidirectional scatter distribution function (BSD F), which is a superset and generalization of both the bidirectional reflectance distribution function (BRDF) and the bidirectional transmission distribution function (BTDF). By analyzing the BSDF, the system captures the scattered light signal over a multitude of directions or angles, potentially covering the entirety of both hemispheres around the object. These measurements can be used to derive detailed information about the shape, size distribution, density, and other properties of the particles, providing a robust characterization framework that supports high-precision analysis and identification.

[0027] B. Spectral Scattering Measurements

[0028]

[0019] The system also incorporates the capability to perform spectral scattering measurements, capturing how the light from the object and the associated BSDF vary as a function of wavelength. This spectral characterization of the objects and particles can span portions of the ultraviolet, visible, and infrared electromagnetic spectrum, allowing the system to capture detailed optical signatures across a broad range of wavelengths. By analyzing the wavelength dependence of scattered light, the system can gain additional insights into the properties of the particles, including structural details and any wavelength-specific scattering behaviors. These spectral analyses enhance the system's ability to perform detailed and nuanced characterizations of the particles, which is crucial for accurately identifying and classifying contaminants and other objects within the fluid. The spectral scattering measurements provide a powerful tool for distinguishing between different types of particles based on their unique optical properties, thereby improving the overall reliability and effectiveness of the system.

[0029] C. Characterization Utilizing Optical Albedo Measurements Over Time

[0030]

[0020] The system is also capable of measuring albedo as a function of time, providing dynamic and temporal information for the recorded optical signals and insights about the size and shape of particles as they move within the microfluidic channel. By tracking changes in reflectivity, transmission, and / or scatter over time, the system can analyze the motion and rotational dynamics of the particles, capturing how their albedo varies as they translate or rotate within the channel. These temporal data offer additional characterization of objects and particles flowing through the channel and opportunities to capture transient phenomena or transitions that may occur over time. Furthermore, by correlating albedo changes with the motion dynamics, the system can derive more detailed physical properties and behaviors, enabling more accurate and dynamic characterizations of both active and inactive particles.

[0031] D. Fluorescent Optical Signal Measurements

[0032]

[0021] Optical characterization of the objects or particles can include advanced capabilities that measure fluorescent optical signals emitted by particles within the channel, offering an additional means of characterization and discrimination based on unique fluorescence properties. By exciting particles with specific wavelengths of light and measuring the emitted fluorescence, the system can identify different types of particles, objects, or materials based on their distinct fluorescence signatures. Fluorescence measurements provide a highly sensitive and specific method for detecting and classifying particles, as different materials and biological entities exhibit unique fluorescence behaviors when exposed to excitation light. This technique enhances the system's ability to detect contaminants and other objects within the fluid, improving the overall accuracy and reliability of the sterility testing process. The incorporation of fluorescence measurements allows for the detection of a wide range of particle types, including microbial contaminants, biological cells, and synthetic materials, making the system versatile and adaptable to various applications.

[0033] E. Raman Scattering Optical Signal Measurements

[0034]

[0022] Other system implementations could incorporate characterization of the objects or particles based on advanced measurement collections of Raman scattering optical signals, providing an additional method for characterizing and discriminating particles. Raman scattering reveals molecular and structural information unique to different materials. By analyzing the Raman spectra, the system can identify various types of particles and contaminants based on their specific Raman signatures. This technique offers a powerful and non-destructive means of obtaining detailed molecular information, enhancing the system's ability to accurately identify and classify particles within the fluid. The Raman scattering measurements complement other optical characterization methods, providing a comprehensive framework for analyzing the composition, structure, and properties of particles.

[0035]

[0023] These enhancements (A-E) or others may be employed to ensure the system provides comprehensive, accurate, and detailed characterizations that improve sterility testing and other liquid quality testing applications through measurements of individual contaminants, particles, or other objects in the fluid.

[0036]

[0024] Utilizing a combination of optical detection and interrogations methods, the system collects and processes multiple components of light information to construct a multifaceted picture of each object. Object classification and identification proceeds by ( / ) a direct approach, using automated physical analysis of the output data components, ( / / ) an artificial intelligence approach, using machine learning, or (Hi) a combined approach.

[0037]

[0025] Evaluation will determine that a solution is sterile unless at least one microbial cell or colony forming unit resides in a 5 mL sample volume (in accordance with the United States Pharmacopeia, USP<71>). Additional results include automatic particle counts relevant to other quality assurance requirements (USP<788>).

[0038] Additional Strategies for Enhancing Volume Flow Rates and Testing Throughput

[0039]

[0026] The design of the system includes several options to control and enhance the total volume of fluid analyzed, aiming to efficiently increase and optimize testing throughput. Key design parameters include fluid channel dimensions, flow rates, and the number of separate, joined, or parallel analyzing channels. By optimizing these parameters, the system can handle larger volumes of fluid within shorter time frames. One straightforward approach to achieve this is through modular parallelization. Integrating multiple CSI modules allows the fluid to be split into several analysis stations, boosting throughput. Incorporating an array of parallel microfluidic channels equipped with individual optical interrogation units multiplies the analysis capacity directly proportional to the number of channels. The overall system can potentially be miniaturized or maintain a small form factor by utilizing micro-optics, forming a compact array of modules for parallel or serial fluid analysis.

[0040]

[0027] Additionally, the system design can address the need to view the entire fluid depth as well as the inherent limitations in the depth of field of microscope objectives. One innovative implementation approach is to hold the fluidic channel at an angle relative to the microscope objective. This configuration allows the objective's in-focus plane to span various depths along the flow direction of the fluidic channel, enabling use of flow channels potentially deeper than upper limits elsewhere specified, and potentially achieving volume flow rates exceeding throughput ranges elsewhere specified. In such an implementation the corresponding in focus z-depth portion of the fluid channel would be allowed to vary spatially across the different associated portions of the image sensor plane. Utilizing such an implementation, a tilted channel approach ensures that particles at different depths can be imaged and analyzed effectively, thereby enhancing the system's ability to process larger fluid volumes rapidly and accurately.

[0041]

[0028] Alternatively, specialized amplitude and / or phase optical components can be integrated into the system to extend the depth of focus of the optical objective(s) viewing the channel to achieve the same goal of interrogating particles over a larger volume with an associated increased system throughput. Preferred methods may include rapidly varying or sweeping through an extended focal range or using a depth extending phase plate conjugate to the pupil plane of the objective. In addition to these implementations, other methods could be chosen to extend the depth of focus beyond capabilities of conventional optical systems.

[0042] Interface for Fluidic Delivery into the System

[0043]

[0029] The system features a versatile fluidic delivery interface to accommodate various sample introduction methods, ensuring seamless integration with diverse fluid sources. This interface allows for the injection of fluids into the system from syringes, bioreactors, sample vials, cups, and other containers holding biological or environmental samples, such as urine, blood, or water. Compatibility with standard laboratory equipment and practices provides a user-friendly and efficient means of introducing samples into the system for analysis. The interface may include components such as syringe ports, luer lock connectors, and adaptable inlet ports that can securely accommodate different container types. Additionally, the interface is designed to minimize sample contamination and ensure accurate volume control, facilitating precise and reliable fluid delivery. The incorporation of automated fluid handling mechanisms can further enhance the efficiency of the sample introduction process, allowing for high-throughput analysis in both laboratory and field settings. This fluidic delivery interface ensures that the system can be easily integrated into existing workflows, providing a robust and flexible solution for various applications in sterility testing, biopharmaceutical quality control, and environmental monitoring.

[0044]

[0030] Any selection of the described features and capabilities can be combined in various embodiments to realize benefits of the CSI system or approach. We have successfully built and tested different CSI hardware configurations and devices, all within the scope of this invention. The instances detailed herein are illustrative rather than limiting. This invention encompasses both simple and advanced versions of the platform for detecting and interrogating various contaminants or subvisible solution borne particles to reveal their type or identity.

[0045] CSI Illustrated in a Second Specific Device

[0046]

[0031] A second embodiment extends the capabilities of the above apparatus and illustrates use of several inspection modes. To enhance the ability to automatically distinguish particulates by type, in this instance, widefield detection throughout the entire channel width and depth is followed by interrogation within a localized volume surrounding the detected object of interest. This proceeds in sequence:

[0047] (1). Detection during rapid fluid flow. On the darkfield camera, the detected object appears as a streak - a length of nonzero pixels proportional to the linear object speed during a camera exposure time. The exposure time and flow rate are tuned so that the streak crosses the sensor without truncating, thereby maximizing sensor performance and flow rate. The total streak intensity is proportional to object size and independent of object speed.

[0048] (2). Stopping and steering. The object is centered on the detector by optical feedback and microfluidic control.

[0049] (3). Switching: detector to interrogator. The flow channel is translated in the direction parallel to its width to move the object from the center of the detector, viewing a wide field, to the center of the interrogator, viewing a local field.

[0050] (4). Focusing. A light sheet - uniform across the local field (0.1 mm x 0.1 mm) - sweeps through the full channel depth (0.1 mm) to find and focus on the object.

[0051] (5). Interrogation. The system automatically records spectrographic data and 3D centroid tracking data timeseries.

[0052]

[0032] The object is kept centered in the lateral (x), flow (y), and focus (z) directions by following its motion in three dimensions. The light sheet illuminating the object coincides with the focal plane of an imaging spectrometer. Background is further reduced by rejecting out of focus light with an aperture. Any constant residual background is automatically subtracted.

[0053]

[0033] An alternative embodiment utilizes 405 nm (violet or near UV) light, or utilizes light of any other wavelength causing objects to fluoresce. Microbial contaminants are identified in this way because their composition includes endogenous fluorophores that emit visible light when excited with 405 nm light or excitation light at other wavelengths. Measurements show that spectrographic signatures may be different for different types of contaminants.

[0054]

[0034] After an interrogation time of less than 30 seconds, the system switches back to the detector and resumes fluid flow. The interrogation sequence begins again if another object of interest is detected. Records of the detection event (streak data), the object motion trajectory (time,x,y,z), and spectrograph timeseries (time, wavelength, intensity) are cataloged and processed by the system to evaluate sterility of biopharmaceutical or other complex fluid products. This method and apparatus herein define a solution for automatic, real-time, in-process quality control monitoring in biomanufacturing environments or in other contexts involving particle detection and flow through analysis of liquid samples.

[0055] BRIEF DESCRIPTION OF DRAWINGS

[0056]

[0035] FIG. 1 depicts a flow channel for viewing microscopic objects in flow.

[0057]

[0036] FIG. 2 shows example data at the CSI checkpoints.

[0058]

[0037] FIG. 3 shows CSI hardware in the first instance.

[0059]

[0038] FIG. 4 shows modular parallelization of CSI hardware.

[0060]

[0039] FIG. 5 shows CSI hardware in the second instance.

[0061]

[0040] FIG. 6 depicts the offset of the detector and interrogator optics in the second instance.

[0062]

[0041] FIG. 7 depicts switching between the detection and interrogation volumes.

[0063]

[0042] FIG. 8 depicts scanning of a light sheet in the interrogation volume.

[0064]

[0043] FIG. 9 depicts 3D tracking in the interrogation volume.

[0065]

[0044] FIG. 10 shows example spectrograph timeseries data.

[0066]

[0045] FIG. 11 shows discernment of two contaminant species by their spectrographic data.

[0067] DETAILED DESCRIPTION

[0068]

[0046] Controlled laminar flow acts as a "conveyor belt" for moving subvisible particles to the machine vision checkpoints of CSI (Figure 1). The subvisible particle or contaminant (101) could be a colony forming unit (CFU) or any other cell or particle type. The fluid channel is wide (102) to carry subvisible particles along many parallel flow paths and increase testing throughput and is shallow (103) to detect particles at any depth. The transparent window for viewing subvisible particles is thickened (104) to keep the outside surface defocused whereas the much shallower fluid depth is in focus.

[0047] The machine vision checkpoints of CSI (Figure 2) capture different types of light-based information. The example shows CSI data on a single colony forming unit residing in a complex fluid (a DNA based biologic). From left to right, the panels show a streak detection event during rapid fluid flow (201) at Checkpoint 1; subpixel tracking with flow stopped (202) at Checkpoint 2; and imaging spectroscopy (203) at Checkpoint 3.

[0069]

[0048] A first instance of CSI hardware (Figure 3) has a rugged and compact design configuration. For scale, a 50 mm length is depicted at the location of the flow channel (301); the flow channel could be 50 mm or any length satisfying laminar Poiseuille flow at a desired flow rate, channel depth, and channel width combination, and has the features depicted in Fig. 1. The apparatus has a diminutive form factor approximately the size of an adult human hand and would be fully enclosed within a lunchbox sized container.

[0070]

[0049] An alternative embodiment uses a continuous broadband LED light source (302) and other light sources could be used. Light signals are collected by optics shown at 0.25 numerical aperture (303) and low magnification designed to view the full channel width and depth. Triple redundant optical detectors include a darkfield camera (304) at 6x magnification, an event camera (305) at 6x magnification, and a spectrometer (306) at 2x magnification.

[0071]

[0050] Modular parallelization of CSI hardware (Figure 4) illustrates a strategy to incrementally enhance throughput or perform multiple measurements serially or simultaneously. For example, 5x stacking of hardware modules (401) could facilitate simultaneous testing from five different locations of a bioreactor using identical CSI modules; or alternatively, a single bioreactor sample could be measured by the 5x series of differently configured CSI modules. The rearmost module also depicts how one or multiple light sources (402) can be strategically positioned to capture forward or backscattering signals of particles.

[0072]

[0051] A second instance of CSI hardware (Figure 5) expands the capabilities of the Fig. 3 design. This instrument uses a detector (501) and an interrogator (502) built to optimize these functions on two different optical paths. The detector is built with a broadband LED light source (503) as in the Fig. 3 design, and in this example uses a redundant darkfield camera (504) plus event camera (505) combination at 0.25 numerical aperture and 3x magnification. Whereas the detector views the entire channel width and depth to count particles in rapid flow, the interrogator uses 3D localization to isolate a single particle anywhere in the detection volume.

[0073]

[0052] The interrogator has an excitation light path (506) and a spectrometer light path (507). 3D localization is achieved with an x-axis for translation of the particle by moving the flow channel (508), a y-axis for fluid steering of the particle (509), and a z-axis for depth scanning of an excitation light sheet and maintaining the particle in focus (510) using an electrically tunable lens.

[0074]

[0053] The optical axes are laterally offset (Figure 6) to preserve darkfield detector conditions without reflections from interrogator optics. The interrogator at 0.75 numerical aperture (601) is proximate to the flow channel, whereas the detector at 0.25 numerical aperture (602) has a longer working distance. Arrows in the x-axis translation direction (603) depict the switching operation.

[0075]

[0054] Switching the object between the widefield detector and localized interrogator is achieved by translating the flow channel between their laterally offset optical axes. An object detected in flow is stopped and then translated to localize the object.

[0076]

[0055] Reversible switching is shown schematically (Figure 7). As in Fig. 6, this operation is depicted with arrows along the x-axis (701). The interrogation volume (702) is much smaller than the detection volume (703).

[0077]

[0056] The object is localized within the interrogation volume (Figure 8). The object (801) is any subvisible particle, cell, or contaminant. The interrogation volume (802) is in the depicted wireframe. A depth scanning light sheet (803) focuses on the object anywhere within the interrogation volume.

[0078]

[0057] The object is tracked in 3D within the interrogation volume (Figure 9). As in Fig. 8, the object (901) is any subvisible particle, cell, or contaminant and the interrogation volume (902) is in the depicted wireframe. The space curve (903) depicts the motion trajectory of the object (time,x,y,z).

[0079]

[0058] Object spectrographs are also recorded by the interrogator (Figure 10). Spectrographic timeseries data (time, wavelength, intensity) are recorded to ascertain object composition. Example data are autofluorescence spectrographs of a microbial contaminant (1001).

[0080]

[0059] Measurements show compositional differences distinguishing two contaminant species (Figure 11). Measurements of each individual species are shown as fluorescence intensity vs wavelength plots (1101). A common reference curve was fit to both plots and subtracted to obtain the residuals distinguishing species (1102). The dashed line is a common reference curve used for comparison.

Claims

Claims1. A system for evaluating the particulate matter contents of liquid samples using a combination of microfluidics and light-based measurements, wherein the system automatically performs single particle measurements throughout the entirety of the sample in a flow through device, and the system further includes hardware and software for detecting, tracking, and recording single particle data either particle by particle or for multiple particles simultaneously a. wherein the detection of particles occurs across the full width and throughout the depth of the microfluidic channel, b. particles to be tracked are localized by hardware or software at any position in the width and depth of the channel, and c. wherein laminar flow carries particles and controls their position in the fluid flow direction.

2. The system of claim 1, wherein the microfluidic channel is preferably wider than it is deep, with a larger cross sectional aspect ratio facilitating higher throughput of liquid volume per unit time without increase to the average particle speed in the flow direction, and wherein multiple parallel particle paths defined by laminar flow can be measured simultaneously by optical sensors.

3. The system of claim 1, wherein one or more sensors capture light information from each particle at a multitude of wavelengths, and preferably across a continuous range of wavelengths, wherein a particle can be viewed as either a localized light image, or as a collection of localized light images distributed by wavelength at UV, visible, or infrared wavelengths.

4. The system of claim 3, wherein optical sensors capture at least one light data type, comprising scattered light, absorbed light, transmitted light, reflected light, or light emitted in a fluorescence process.

5. The system of claim 4, wherein the light signal is attributable to the particle or its interface with the surrounding fluid, without altering the fluid or particle composition.

6. The system of claim 4, wherein the fluid composition is altered to change the contrast of particles.

7. The system of claim 4, wherein the particle composition is altered to enhance the contrast of particles using either chemically specific or nonspecific contrast labels.

8. The system of claim 1, wherein the optical components and associated sensors comprise a single light source or multiple light sources positioned to illuminate objects or particles in the fluid from various directions, enabling measurements of subsets or the entire total integrated scatter (TIS) and bidirectional scatter distribution function (BSDF) to derive particle information such as shape, size, and density information.

9. The system of claim 8, wherein the light sources include at least one overhead top-down illumination and one bottom-up illumination, providing forward and reverse directional characterizations of the optical properties of the particles.

10. The system of claim 1, wherein modular parallelization of any single instance of the device or multiple device instances facilitates testing of multiple fluid samples simultaneously, or the same fluid sample serially, and wherein samples may be obtained from one or multiple sources or multiple locations of the same source, including but not limited to different locations of a bioreactor or process path, wherein the throughput and volume flow rate of fluid analyzed may be increased by splitting the fluid or fluids into multiple analysis stations, each equipped with individual optical interrogation units to multiply analysis capacity.

11. The system of claim 1, wherein the fluidic channel is held at an angle relative to the microscope objective, allowing the in-focus plane of the objective to span various depths along the channel in the flow direction, enabling detection of particles at different depths of a channel that may be deeper than specified elsewhere in this document, and that thereby facilitate testing of fluids at higher volume flowrates than specified elsewhere in this document.

12. The system of claim 1, wherein spatialized amplitude and / or phase optics are incorporated to extend the depth of field of microscope objectives to cover an extended depth of the fluidic channel with an associated increase in throughput and volume flow rates.

13. The system of claims 1, wherein variations of the collected scattered light as a function of wavelength are used for the characterization and determination of particles in a fluid, and wherein the scattering data of particles relate to their structure, material composition, or other aspects of physically interpretable scattering processes.

14. The system of claim 1, wherein dynamic optical characterization of objects or particles in the fluid is achieved by measuring albedo as a function of time, providing information on particle size, shape, motion dynamics, and transient phenomena of the particles.

15. The system of claim 1, wherein the optical characterization of particles is conducted by measurement of Raman scattering optical signals that provide compositional information and molecular structure details for particle identification, and wherein such data are analyzed to discriminate between different types of particles and contaminants based on their unique Raman signatures.

16. The device of claim 1, wherein at least one sensor operating as a widefield particle detector captures light at a numerical aperture between 0.05 and 0.3, and at a magnification less than lOx, and in which an optimal numerical aperture and magnification are chosen together to maximize light gathering efficiency on the sensor while maintaining a field of view across the entire fluid channel width and depth, and wherein the microfluidic channel has a material viewing window thicker than the channel depth, causing the outside of the channel to be defocused whereas fluid and particles inside the channel are in focus.

17. The device of claim 1 that is miniature and rugged and that has triple redundant particle detectors comprising a photon counting camera or neuromorphic event camera, a video camera, and a spectrometer and these simultaneously or serially detect particles throughout the width and depth of the fluidic channel without need of moving parts.

18. The device of claim 1 in a second preferred embodiment that switches from a widefield particle detection mode to a localized particle interrogation mode, wherein the interrogation mode is triggered by detection of a particle of interest within a detection volume of at least 0.1 cubic millimeters or 100 nanoliters, and the particle is then localized to a smaller measurement volume, preferably no larger than 0.001 cubic millimeters or 1 nanoliter, and wherein a particle can be localized for interrogation at any point of the detection volume.

19. The device of claim 18, wherein the autofluorescence of a particle is measured on a spectrometer to ascertain its composition.

20. The device of claim 19, wherein the excitation light is confined to a region no larger than the interrogation volume traversed by a depth scanning light sheet, and preferably confined to a minimum traversed volume wherein the particle is in focus, thereby maximizing signal to noiseof fluorescence spectroscopy measurements and minimizing phototoxicity to the sample by confining the excitation volume to the optical section selectively illuminating the particle under interrogation.

21. The system of claim 1, wherein a particle in flow is detected on a photon counting or neuromorphic event camera.

22. The system of claim 1, wherein a particle in flow is detected on an imaging camera as a streak, with streak length defined in a frame exposure time by a series of nonzero pixel values in the direction of motion.

23. The system of claim 1, wherein a particle in flow is detected by nonzero values on a spectrometer.

24. The system of claim 1, wherein the position of a particle is measured in one, two, or three dimensions to ascertain its motion trajectory, and wherein one or more of particle size, solution viscosity, or possible motility parameters are measured.

25. The system of claim 1, wherein a spectrograph time series is recorded to measure time evolution of fluorescence processes or structural fluctuations of particles within light scattering processes.

26. The system of claim 1, wherein the scattered light signal of a particle is captured to ascertain its structure, including estimates of particle size and particle shape, and the refractive index change at the particle / fluid interface.

27. The system of claim 1, wherein measurements related to the composition, motion, and structure of each particle are collected in the form of records of each particle captured in a sparse data format, wherein the records include- spectrographic intensity vs wavelength profiles at multiple timepoints,- position vs time trajectory data, and- the scattered light intensity data as described in claims 14.

28. The system of claim 27, wherein the position vs. time trajectory data comprise measurements of the position of a particle measured in one, two, or three dimensions to ascertain its motion trajectory, and wherein one or more of particle size, solution viscosity, or possible motility parameters are measured.

29. The system of claim 27, wherein the spectrographic intensity vs. wavelength profiles at multiple timepoints is in the form of a position vs. a spectrograph time series which is recorded to measure time evolution of fluorescence processes or structural fluctuations of particles within light scattering processes.

30. The system of claim 27, wherein the scattered light intensity data comprises a signal of a particle captured to ascertain its structure, including estimates of particle size and particle shape, and the refractive index change at the particle / fluid interface.

31. The system of claim 27, wherein the scattered light intensity data comprises dynamic optical characterization of objects or particles in the fluid achieved by measuring albedo as a function of time, thereby providing information on particle size, shape, motion dynamics, and transient phenomena of the particles.

32. The system of claim 1, further comprising a software system for classifying particles based on measured attributes, including structural, motility, and compositional parameters.

33. The system of claim 1, further comprising a software system which employs machine learning or artificial intelligence to classify particles.

34. A system according to claim 32 or 33, wherein liquid samples are automatically evaluated and determined to be sterile or nonsterile, and in which subvisible particles of any type are classified and counted.

35. The system of claim 1, wherein the fluidic delivery interface for introducing samples into a microfluidic analysis system is configured to adapt to a variety of settings, including both laboratory and field settings that may utilize syringe ports, luer lock connectors, or other interfaces that the system accommodates with adaptable inlet ports designed to securely accommodate syringes, bioreactors, sample vials, cups, and other containers holding biological or environmental samples.

36. The fluidic delivery interface of claim 35, further comprising automatic fluid handling mechanisms which enhance efficiency, minimize contamination, and ensure precise volume control during sample introduction into the system.

37. A system according to claim 1, wherein the liquid is water.

38. A system according to claim 1, wherein the liquid is a bodily fluid including but not limited to blood, urine, or saliva.

39. A system according to claim 1, wherein the liquid is a medicine, including but not limited to a biologic or biopharmaceutical product.

Images