Integrated microfluidic system for the generation of microorganospheres (MOS)

Abstract

The microfluidic device includes a microfluidic chip for generating MicroOrganoSpheres (MOS). A first channel is defined on the surface of the microfluidic chip and includes a droplet generation portion, the droplet generation portion including an inlet portion, a junction between the inlet portion and an emulsification fluid channel, and a chamber downstream of the junction. The cross-sectional area of ​​the chamber is larger than the cross-sectional area of ​​the inlet portion. The first channel includes a polymerization portion downstream of the droplet generation portion, the polymerization portion having a serpentine shape. The device includes a cartridge for MOS demulsification, the cartridge including a collection reservoir, a substrate disposed on the collection reservoir, and a membrane disposed between the collection reservoir and the surface of the substrate. A second channel is defined on the surface of the substrate facing the collection reservoir and is fluidly connected to the outlet of the polymerization portion of the first channel.

Description

[Technical Field]

[0001] Priority claims This application claims the benefit of priority to U.S. Patent Application No. 63 / 415,228, filed October 11, 2022, U.S. Patent Application No. 63 / 415,240, filed October 11, 2022, U.S. Patent Application No. 63 / 415,232, filed October 11, 2022, and U.S. Patent Application No. 63 / 415,235, filed October 11, 2022, each of which is incorporated by reference in its entirety herein. [Background technology]

[0002] Patient-derived biological material, such as cells obtained from biopsy or excised tissue, can be used to screen for treatments that the patient will respond to effectively. Screening for treatments can be done using microfluidic devices. Summary of the Invention

[0003] In a first aspect, a microfluidic device includes: a microfluidic chip for producing MicroOrganoSpheres (MOS), the microfluidic chip having a first microfluidic channel defined on a surface thereof, the first microfluidic channel including an inlet portion, a junction between the inlet portion and an emulsification fluid channel, and a droplet generation portion including a chamber downstream of the junction, the droplet generation portion having a cross-sectional area larger than that of the inlet portion, and a polymerization portion downstream of the droplet generation portion, the polymerization portion having a serpentine shape; and a cartridge for MOS demulsification, the cartridge including a collection reservoir, a substrate disposed on the collection reservoir, the substrate having a second microfluidic channel defined on a surface of the substrate facing the collection reservoir, the second microfluidic channel fluidly connected to an outlet of the polymerization portion of the first microfluidic channel, and a membrane disposed between the collection reservoir and the surface of the substrate.

[0004] Embodiments may include one or any combination of two or more of the following features.

[0005] The droplet generation portion of the first microfluidic channel includes an outlet portion downstream of a chamber, the cross-sectional area of ​​the chamber being larger than the cross-sectional area of ​​the outlet portion, and in some cases, at least some of the outlet portions extend in a direction parallel to the chamber.

[0006] The surface of the microfluidic chip is a first surface, and a polymerized portion of the microfluidic channel is defined on the first surface of the microfluidic chip and on a second surface of the microfluidic chip opposite the first surface.

[0007] The junction comprises a junction of two hydrophobic fluidic channels. In some cases, the junction is a right-angle junction.

[0008] The membrane includes a hydrophobic membrane. In some cases, the membrane is both hydrophobic and lipophilic.

[0009] The second microfluidic channel includes an upstream section with a simple serpentine shape and a downstream section with a double serpentine shape.

[0010] The cross-sectional area of ​​the second microfluidic channel decreases from the inlet end of the second microfluidic channel to the outlet end of the second microfluidic channel.

[0011] The surface of the substrate is a first surface, and a media inlet channel is defined on a second surface of the substrate opposite the first surface of the substrate, the media inlet channel being fluidly connected to an upstream section of the second microfluidic channel and configured to be connected to a media reservoir. In some cases, the demulsification cartridge includes a media reservoir. In some cases, the media inlet channel is fluidly connected to the media reservoir via a tube extending through the substrate and the collection reservoir. In some cases, a collection reservoir is disposed within a cavity defined in the media reservoir such that the collection reservoir is positioned between the media reservoir and the substrate. In some cases, a bottom surface of the media reservoir is angled with respect to the plane of the substrate. In some cases, the demulsification cartridge includes a duckbill valve extending through the substrate and the collection reservoir, the duckbill valve being configured to provide fluid access to the media reservoir.

[0012] The demulsifying cartridge includes a hydrophobic material disposed within a collection reservoir.

[0013] A vacuum channel is defined through the body of the collection vessel and is configured to allow a vacuum to be applied to a surface of the membrane opposite the substrate.

[0014] The microfluidic device includes a reservoir fluidly connected to a first microfluidic channel through an inlet port defined at an inlet end of the first microfluidic channel. In some cases, the reservoir includes a base and a cover, the base and the cover defining a cavity for a fluid sample. In some cases, the microfluidic device includes an inlet port in the reservoir cover, the inlet port including a duckbill valve. In some cases, the microfluidic device includes an outlet port in the reservoir cover, the outlet port connected to a tube extending into the reservoir cavity. In some cases, a bottom surface of the reservoir base is angled relative to the cover. In some cases, the microfluidic device includes a reservoir holder configured to receive the reservoir, the reservoir holder including a cooling system configured to cool the reservoir. In some cases, the cooling system includes a thermoelectric cooling system.

[0015] One or more cutouts are defined between the droplet generation portion and the polymerization portion of the microfluidic chip. In some cases, edges of the one or more cutouts are angled relative to a surface of the microfluidic chip. In some cases, the one or more cutouts extend through the entire thickness of the microfluidic chip.

[0016] The microfluidic device includes a cover disposed on a surface of the microfluidic chip. In some cases, the cover includes an optically transparent cover.

[0017] A plurality of first microfluidic channels are defined on a surface of the microfluidic chip, and the device includes a plurality of cartridges, each of the second microfluidic channels of the cartridges being fluidly connected to a corresponding one of the first microfluidic channels of the microfluidic chip.

[0018] The device includes an effluent vial fluidly connected to the second microfluidic channel through an outlet port defined in the outlet end of the second microfluidic channel.

[0019] In a second aspect that may be combined with any embodiment of the preceding aspects, a system includes the microfluidic device of the first aspect; a housing, wherein the microfluidic device is disposed within the housing; and a polymerization block contained within the housing and positioned to apply a stimulus to a polymerization portion of the first microfluidic channel.

[0020] Embodiments may include one or any combination of two or more of the following features.

[0021] The polymerization block includes a thermal polymerization block configured to apply heat to a polymerization portion of the first microfluidic channel. Optionally, the thermal polymerization block includes a heater. Optionally, the thermal polymerization block includes a temperature sensor. Optionally, the temperature sensor includes one or more of a thermistor, a thermocouple, or a resistance temperature detector. Optionally, the system includes a controller configured to control operation of the resistive heater in response to temperature data received from the temperature sensor. Optionally, the heater includes a resistive heater. Optionally, the thermal polymerization block includes an insulating cover, with the heater disposed within a cavity defined within the insulating cover.

[0022] The polymerization block includes a photopolymerization block configured to illuminate a polymerization portion of the first microfluidic channel. Optionally, the photopolymerization block includes a light-emitting diode (LED). Optionally, the photopolymerization block includes a photodetector. Optionally, the system includes a controller configured to control operation of the LED in response to light intensity data received from the photodetector. Optionally, the LED is disposed within a cavity defined within a housing of the photopolymerization block. Optionally, a wall of the cavity is formed from a material capable of reflecting light at a wavelength of light output by the LED. Optionally, the system includes a controller configured to control the LED to emit pulsed illumination.

[0023] The surface of the microfluidic chip is a first surface, and the polymerized block includes a first block disposed adjacent to the first surface of the microfluidic chip and a second block disposed adjacent to a second surface of the microfluidic chip, the second surface being opposite the first surface. In some cases, the first block and the second block are fixed relative to the microfluidic chip by a spring. In some cases, the first block and the second block are clamped to the microfluidic chip.

[0024] The system includes a reservoir for emulsifying fluid, and an emulsifying fluid channel of the microfluidic device is fluidly connected to the reservoir. Optionally, the reservoir includes a reflective rib for fluid volume measurement disposed within a chamber of the reservoir. Optionally, the system includes a pump disposed between the reservoir for emulsifying fluid and the emulsifying fluid channel. Optionally, the system includes a controller configured to control operation of the pump. Optionally, the controller is configured to control operation of the pump to achieve a target fluid velocity in the second microfluidic channel. Optionally, the pump is a syringe pump. In some embodiments, a valve, such as a servo valve, may be used instead of the pump.

[0025] The system includes an imaging system positioned to capture images of at least a portion of the chamber. Optionally, the system includes a controller configured to control a flow rate of a fluid through an inlet portion of the microfluidic channel based on images captured by the imaging system. Optionally, the controller is configured to control the flow rate of the fluid by controlling a pressure applied to a reservoir fluidly connected to the inlet portion of the microfluidic channel. Optionally, the controller is configured to control the flow rate of the fluid by controlling a syringe pump. Optionally, the flow rate of the sample-containing fluid is controlled by pressure, and the flow rate of the emulsifying fluid, such as oil, is controlled by a syringe pump.

[0026] In a third aspect that may be combined with any embodiment of either or both of the preceding aspects, a microfluidic chip includes a plurality of first microfluidic channels for generating an emulsion of droplets of a first fluid in a second fluid, the first microfluidic channels defined in a first surface of the microfluidic chip, each first microfluidic channel being fluidically independent from each other first microfluidic channel, each first microfluidic channel including an inlet portion configured to receive the first fluid from a respective source of the first fluid, a junction between the inlet portion and a corresponding second fluid channel configured to carry the second fluid; and a chamber downstream of the junction, the cross-sectional area of ​​the chamber being greater than the cross-sectional area of ​​the inlet portion; and a plurality of second microfluidic channels for polymerizing droplets of the emulsion, thereby generating a MOS, each second microfluidic channel being fluidly connected to an outlet of a corresponding one of the first microfluidic channels, and each second microfluidic channel being a serpentine channel including a first portion defined on a first surface of the microfluidic chip and a second portion defined on a second surface of the microfluidic chip opposite the first surface.

[0027] Embodiments may include one or any combination of two or more of the following features.

[0028] Each first microfluidic channel includes an outlet portion downstream of a chamber, the cross-sectional area of ​​the chamber being larger than the cross-sectional area of ​​the outlet portion. In some cases, a region of the outlet portion of each first microfluidic channel extends in a direction parallel to the respective chamber.

[0029] The microfluidic chip includes a cover disposed on each of the first surface and the second surface of the microfluidic chip. In some cases, the cover includes an optically transparent cover.

[0030] A plurality of first microfluidic channels are defined in a first region of the microfluidic chip, and a plurality of second microfluidic channels are defined in a second region of the microfluidic chip different from the first region. In some cases, one or more cutouts are defined between the first and second regions of the microfluidic chip. In some cases, edges of the one or more cutouts are angled relative to the first and second surfaces of the microfluidic chip. In some cases, the one or more cutouts extend through the entire thickness of the microfluidic chip.

[0031] Each junction is between a respective inlet portion and two corresponding second fluid channels. In some cases, the junction is a right-angle junction.

[0032] The microfluidic chip includes a plurality of inlet fingers, each inlet finger extending away from at least one other inlet finger and separated from each adjacent inlet finger by a gap, and at least some of the inlet portions of each first microfluidic channel defined on the surface of the corresponding inlet finger.

[0033] The microfluidic chip includes a plurality of outlet fingers, each of which extends away from at least one other outlet finger and is separated from each adjacent outlet finger by a gap, and an outlet portion of each second microfluidic channel is defined on a surface of a corresponding outlet finger.

[0034] The outlet port of each second microfluidic channel is configured to be connected to a corresponding cartridge for demulsification of the emulsion.

[0035] In a fourth aspect that can be combined with any one or more of the preceding aspects, an apparatus includes a cartridge for transferring MOS from an emulsion in a hydrophobic fluid to a suspension in an aqueous fluid, e.g., a hydrophilic aqueous fluid such as a growth medium, the demulsifying cartridge including: a collection reservoir defining a cavity for receiving the hydrophobic fluid; a substrate disposed on the collection reservoir, the substrate having a microfluidic channel defined in a first surface of the substrate facing the collection reservoir, the substrate having a medium inlet channel for the aqueous fluid fluidly connected to an upstream portion of the microfluidic channel; and a hydrophobic membrane disposed between the collection reservoir and the surface of the substrate. In some cases, the hydrophobic membrane is both hydrophobic and lipophilic.

[0036] Embodiments may include one or any combination of two or more of the following features.

[0037] The device includes a media reservoir having a cavity configured to contain an aqueous fluid, and the media inlet channel is fluidly connected to the media reservoir. In some cases, the device includes a tube extending through the substrate and the collection container, and the media inlet channel is fluidly connected to the media reservoir via the tube. In some cases, the collection container is disposed within the cavity of the media reservoir such that the collection container is positioned between the media reservoir and the substrate. In some cases, the bottom surface of the media reservoir is angled with respect to the plane of the substrate. In some cases, the device includes a duckbill valve disposed through an opening in the substrate and an opening in the collection container, the duckbill valve configured to allow the aqueous fluid to be provided to the cavity of the media reservoir but prevent backflow.

[0038] The surface of the substrate is a first surface, and a media inlet channel is defined on a second surface of the substrate opposite the first surface.

[0039] The cross-sectional area of ​​the microfluidic channel is larger at the upstream end of the microfluidic channel than at the downstream end of the microfluidic channel.

[0040] The upstream portion of the microfluidic channel has a different shape than the downstream portion of the microfluidic channel. In some cases, the upstream portion of the microfluidic channel has a simple serpentine shape and the downstream portion of the microfluidic channel has a double serpentine shape.

[0041] The device includes a hydrophobic absorbent material disposed within a cavity of a collection reservoir. In some cases, the device includes a material that is both hydrophobic and oleophilic.

[0042] In a fifth aspect that can be combined with any one or more embodiments of the preceding aspects, a method includes: generating droplets of a first fluid in a hydrophobic fluid in a droplet generation portion of a first microfluidic channel defined in a surface of a microfluidic chip, the first fluid comprising a biomaterial and a matrix material; applying a stimulus to the generated droplets in a polymerization portion of the first microfluidic channel to polymerize the matrix material, thereby forming MOS emulsified in the hydrophobic fluid; and transferring the MOS from the emulsion to a suspension in an aqueous fluid, the transfer comprising flowing the mixture of the aqueous fluid and the emulsion of MOS in the hydrophobic fluid along a second microfluidic channel defined in the substrate; and transferring the hydrophobic fluid through a membrane forming a wall of the second microfluidic channel as the mixture flows along the second microfluidic channel.

[0043] Embodiments may include one or any combination of two or more of the following features.

[0044] Generating droplets of the first fluid includes generating droplets at a junction between the first microfluidic channel and one or more channels carrying a hydrophobic fluid. In some cases, the method includes controlling a flow rate of the hydrophobic fluid.

[0045] The method includes controlling a flow rate of the first fluid based on the determined size of the generated droplets. In some cases, the method includes determining a size of the generated droplets based on an image of the droplets in the droplet generation portion of the first microfluidic channel.

[0046] Stimulating the generated droplets includes heating the droplets.

[0047] Stimulating the generated droplets includes irradiating the droplets with light having a wavelength configured to induce polymerization of the matrix material. In some cases, the surface of the microfluidic chip is a first surface, and the polymerization portion of the first microfluidic channel is defined on both the first surface and a second surface of the microfluidic chip, and irradiating the droplets includes irradiating the first surface and the second surface of the microfluidic chip. In some cases, irradiating the droplets includes irradiating the droplets with pulsed illumination.

[0048] The method includes receiving the hydrophobic fluid to be transferred in a collection reservoir, with a membrane disposed between the collection reservoir and the substrate.

[0049] The hydrophobic fluid is moved through the membrane by a pressure difference between a positive driving pressure on the upper side of the membrane and ambient pressure on the lower side, plus the additional force of gravity. In some cases, moving the hydrophobic fluid through the membrane includes applying a vacuum to the membrane.

[0050] The method includes providing a suspension of MOS in an aqueous fluid to an effluent vial.

[0051] The method includes generating droplets of each of a plurality of first fluids in each of a plurality of fluidically independent first microfluidic channels defined on a surface of a microfluidic chip, and applying a stimulus to the generated droplets in each first microfluidic channel to form a MOS.

[0052] The approach described herein may have one or more of the following advantages: MicroOrganoSpheres (MOS) can be produced and extracted from immiscible production fluids in a fast, high-throughput process that provides high recovery rates; The system is fully automated, providing consistent performance from sample to sample; The system utilizes disposable components in the sample processing pathway, reducing the need for downtime for cleaning and disinfection.

[0053] The details of one or more embodiments are set forth in the accompanying drawings and the description below. Other features and advantages will be apparent from the description and drawings, and from the claims. [Brief explanation of the drawings]

[0054] [Figure 1] FIG. 1 is a block diagram of a system for generating MicroOrganoSpheres (MOS).

[0055] [Figure 2] FIG. 1 is a diagram of consumable parts of a MOS generation system.

[0056] [Figure 3A] FIG. 1 is a diagram of the microfluidic chip of the MOS generation system. [Figure 3B] FIG. 1 is a diagram of the microfluidic chip of the MOS generation system.

[0057] [Figure 4] FIG. 1 is a diagram of a portion of a microfluidic chip of a MOS generation system.

[0058] [Figure 5A] FIG. 1 is a diagram of the heat block of the MOS generation system. [Figure 5B] FIG. 1 is a diagram of the heat block of the MOS generation system. [Figure 5C] FIG. 1 is a diagram of the heat block of the MOS generation system.

[0059] [Figure 6A] FIG. 1 is a diagram of the optical block of a MOS generation system. [Figure 6B] FIG. 1 is a diagram of the optical block of a MOS generation system. [Figure 6C] FIG. 1 is a diagram of the optical block of a MOS generation system.

[0060] [Figure 7A] FIG. 1 is a diagram of an exemplary oil reservoir. [Figure 7B] FIG. 1 is a diagram of an exemplary oil reservoir. [Figure 7C] FIG. 1 is a diagram of an exemplary oil reservoir.

[0061] [Figure 7D] FIG. 1 is a diagram of an exemplary oil reservoir. [Figure 7E] FIG. 1 is a diagram of an exemplary oil reservoir.

[0062] [Figure 8A] FIG. 1 is a diagram of an exemplary oil delivery subsystem.

[0063] [Figure 8B] FIG. 1 is a diagram of an exemplary oil delivery subsystem.

[0064] [Figure 9A] FIG. 2 is a cross-sectional view of a sample reservoir. [Figure 9B] FIG. 1 is a perspective view of a sample reservoir.

[0065] [Figure 10A] FIG. 1 is a diagram of multiple reservoirs. [Figure 10B] FIG. 1 is a diagram of multiple reservoirs.

[0066] [Figure 11A] FIG. 1 is a diagram of a cold block.

[0067] [Figure 11B] FIG. 1 is a diagram of a thermoelectric cooling subsystem.

[0068] [Figure 12] FIG. 1 is a diagram of the sample driving subsystem of the MOS generation system.

[0069] [Figure 13A] FIG. 1 is a diagram of an imaging system for a MOS generation system. [Figure 13B] FIG. 1 is a diagram of an imaging system for a MOS generation system.

[0070] [Figure 14] FIG. 1 is a schematic diagram of a portion of the demulsification subsystem of a MOS production system.

[0071] [Figure 15] FIG. 1 is a cross-sectional view of a portion of a demulsification subsystem.

[0072] [Figure 16A] FIG. 1 is a diagram of the demulsification subsystem of the MOS generation system. [Figure 16B] FIG. 1 is a diagram of the demulsification subsystem of the MOS generation system. [Figure 16C] FIG. 1 is a diagram of the demulsification subsystem of the MOS generation system.

[0073] [Figure 17A] FIG. 1 is a diagram of the microfluidic channel geometry in the demulsification subsystem of the MOS production system. [Figure 17B] FIG. 1 is a diagram of the microfluidic channel geometry in the demulsification subsystem of the MOS production system.

[0074] [Figure 18A] FIG. 1 is a diagram of a loading system. [Figure 18B] FIG. 1 is a diagram of a loading system.

[0075] [Figure 19] 1 is a flowchart.

[0076] [Figure 20] FIG. 1 is a diagram of the electronic subsystem of a MOS generation system.

[0077] [Figure 21] FIG. 1 is a block diagram of a MOS generation system.

[0078] [Figure 22] 1 is a flowchart.

[0079] [Figure 23A] FIG. 1 is a schematic diagram of multiple images of a droplet.

[0080] [Figure 23B] FIG.

[0081] [Figure 24A] FIG. 1 is a diagram of a light source.

[0082] [Figure 24B] FIG.

[0083] [Figure 25] 1 is a flowchart.

[0084] [Figure 26A] FIG. 1 is a block diagram of an exemplary device for generating droplets.

[0085] [Figure 26B] FIG. 1 is a perspective view of an exemplary device for generating droplets.

[0086] [Figure 26C] FIG. 1C is a schematic diagram of a portion of the exemplary device shown in FIG. 1B.

[0087] [Figure 27A] 1 is a captured exemplary image of a droplet generation device. [Figure 27B]1 is a captured exemplary image of a droplet generation device. [Figure 27C] 1 is a captured exemplary image of a droplet generation device. [Figure 27D] 1 is a captured exemplary image of a droplet generation device. [Figure 27E] 1 is a captured exemplary image of a droplet generation device.

[0088] [Figure 28] 1 is a flow chart of a process for controlling the pressure of a fluid stream in a droplet generation device.

[0089] [Figure 29] 1 illustrates a process for estimating droplet size.

[0090] [Figure 30] 1 is a graph showing experimental data. [Figure 31] 1 is a graph showing experimental data.

[0091] [Figure 32] 1 is a flowchart of a process for controlling the size of droplets produced by the device.

[0092] [Figure 33A] 1 illustrates an exemplary container including protrusions for liquid level sensing. [Figure 33B] 1 illustrates an exemplary container including protrusions for liquid level sensing. [Figure 33C] 1 illustrates an exemplary container including protrusions for liquid level sensing.

[0093] [Figure 34A] 33A-33C show an example of a liquid level sensor measuring the liquid level in the container of FIGS. 33A-33C when the liquid level is lower than the light incident position. [Figure 34B]33A-33C show an example of a liquid level sensor measuring the liquid level in the container of FIGS. 33A-33C when the liquid level is higher than the light incident position.

[0094] [Figure 35A] 1 illustrates an exemplary liquid level measurement by moving a light source and photodetector pair along the length of a container.

[0095] [Figure 35B] 35B shows exemplary measurement results using the liquid level measurement of FIG. 35A.

[0096] [Figure 36A] 10 illustrates another exemplary liquid level measurement with multiple light source and light detector pairs positioned at a series of locations along the length of the container.

[0097] [Figure 36B] 36B shows another exemplary measurement result using the liquid level measurement of FIG. 36A.

[0098] [Figure 37] 1 illustrates an exemplary fluid system including at least one liquid level sensor.

[0099] [Figure 38] 1 is a flowchart of an exemplary process for liquid level sensing using optical reflection.

[0100] [Figure 39A] This is an image of a MicroOrganoSphere. [Figure 39B] This is an image of a MicroOrganoSphere. [Figure 39C] This is an image of a MicroOrganoSphere.

[0101] [Figure 40] FIG. 1 illustrates an example computing environment. DETAILED DESCRIPTION OF THE INVENTION

[0102] This disclosure describes an integrated microfluidic channel for generating droplets containing biomaterials, polymerizing the droplets to form MicroOrganoSpheres (MOS), and demulsifying the MOS into a suspension in an aqueous fluid, e.g., a hydrophilic aqueous fluid such as growth medium. Specifically, multiple fluidically independent microfluidic channels are defined on a microfluidic chip. Each channel has a droplet generation region configured to generate an emulsion of droplets in a hydrophobic fluid and a serpentine polymerization region where a stimulus, such as light or heat, is applied to polymerize the matrix material in the droplets, thereby forming MOS. Downstream of the polymerization region, each channel is fluidically connected to a microfluidic channel in a demulsification cartridge. A mixture of the aqueous fluid and the MOS in the hydrophobic fluid flows along the microfluidic channel, and the hydrophobic fluid is drawn through a hydrophobic membrane, leaving the MOS suspended in the aqueous fluid.

[0103] The droplet generation and polymerization and MOS demulsification processes using this system are automated and continuous, allowing for simultaneous but independent processing of multiple samples. The automated MOS generation and demulsification process is efficient and reliable, and less prone to errors that can occur during manual processes. For example, a 100 μL sample of biomaterial can be fully processed in less than 10 minutes, e.g., 6 minutes, and a 1 mL sample can be processed in less than 40 minutes, all with hands-off automated processing times.

[0104] MicroOrganoSpheres (MOS) are generally spherical structures containing biomaterials, such as dissociated tissues, e.g., cells, dispersed (e.g., suspended) in a matrix material. MOSs can be used, for example, for high-throughput patient-specific screening of effective treatments. The biomaterials in MOSs can be microbiopsies, excised tissue (e.g., from tumors or organs), or cells extracted from other sources (e.g., banked or cultured cells). Each MOS contains a small number of cells, e.g., a cluster of 5-10 cells, suspended in the matrix material. The matrix material is a material that can polymerize upon application of a stimulus (e.g., heat, light, or chemical reaction) to form a biomaterial support or support network. For example, the matrix material can be a hydrogel.

[0105] A large number of MOSs can be generated from a single patient sample, e.g., a single biopsy or tissue resection. For example, from a tissue sample (e.g., a needle biopsy sample) having a volume of approximately 10 to 1000 μL, more than 10,000 substantially uniformly sized MOSs (e.g., more than 20,000, more than 30,000, more than 40,000, more than 50,000, more than 60,000, more than 70,000, more than 80,000, more than 90,000, or more than 100,000) can be generated. This large number of MOSs can be used for high-throughput screening of multiple treatments to identify which treatment a particular patient will respond to. For example, MOSs containing cells from a patient biopsy can be used for high-throughput screening of drug compositions that can predict which treatments will be effectively applied to that patient. This screening may, for example, enable efficient screening of multiple potential treatments (e.g., drug treatments) to identify an effective treatment for a cancer patient before treating that patient. The generation of MOSs and their use for screening is further described below and in US2020 / 0377861, the contents of which are incorporated herein by reference in their entirety.

[0106] The harvested MOS can be used, for example, to screen for potential treatments for the patient who provided the biomaterial to the MOS. For example, the harvested MOS can be placed in culture medium to grow cells within the MOS. After culturing, the cells can be assayed substantially immediately or cryopreserved for future use.

[0107] Referring to FIG. 1 , an apparatus for MOS generation 100 contains integrated microfluidic channels that provide three aspects of MOS generation functionality: droplet generation 102, droplet polymerization 104, and demulsification 106. In droplet generation 102, droplets containing patient-specific biomaterial and matrix material (e.g., hydrogel) are generated as an emulsion in a fluid (e.g., oil) that is immiscible with the matrix material. In droplet polymerization 104, the matrix material in the generated droplets is polymerized, thereby forming MOS. In demulsification 106, the MOS is transferred from the oil emulsion to a suspension in an aqueous fluid. The resulting suspension can be used for downstream processing (e.g., further cell growth) and / or testing. The apparatus 100 accepts patient-specific biomaterial 110 (e.g., dissociated tissue such as cells), emulsion oil 112, and aqueous fluid 114 as inputs, and outputs MOS 116 in the aqueous fluid and waste products 118, such as oil from the emulsion. Droplet generation 102 and droplet polymerization 104 are performed on a microfluidic chip that is fluidly connected to a demulsification cartridge for demulsification 106 .

[0108] The device 100 is capable of processing multiple samples, such as four samples (e.g., samples from multiple patients), in parallel (e.g., simultaneously) along multiple fluidically separated flow paths. The portions of the device 100 that contact the samples are disposable so that they do not need to be cleaned or disinfected between the processing of multiple samples.

[0109] Droplet polymerization occurs through cross-linking of a matrix material in which the biomaterial is suspended. The matrix material can be cross-linked by applying a thermal stimulus (e.g., heat), by a photoinitiated process, or by a chemical reaction. The device 100 can be configured to accommodate one or more of these cross-linking approaches. For example, the device 100 can be equipped with hardware for either thermal or photoinitiated polymerization, and can also be adapted for chemical polymerization through the use of a suitable microfluidic chip.

[0110] FIG. 2 illustrates an example of a MOS generation system 200 that can be used, for example, to generate MOS in the instrument 100 of FIG. 1. The MOS generation system 200 is configured for parallel processing of multiple samples provided to its respective reservoirs 220. In the illustrated example, the MOS generation system can process up to four samples simultaneously. Each sample contains a biomaterial (e.g., patient-specific biomaterial, such as tumor cells) suspended in a matrix material, such as a hydrogel. The sample from each reservoir 220 is processed in an independent microfluidic channel (e.g., fluidically isolated and independently controllable) defined in the microfluidic chip 210, where the droplet generation and polymerization process take place. After droplet polymerization, the MOS emulsion in each channel is provided to a corresponding demulsification cartridge 230, which transfers the MOS from oil to a suspension in an aqueous fluid. The MOS in the aqueous fluid is then effluent-discharged into an effluent vial 240.

[0111] To process a set of up to four samples, the MOS generation system 200 is loaded into a MOS generation instrument. Optionally, the desired type of polymerization stimulation is enabled, for example, by inserting a heat block or light block into the instrument for thermal or light-initiated polymerization, respectively. Once processing of a set of samples is complete, the MOS generation system 200 (including the reservoir 220, microfluidic chip 210, demulsification cartridge 230, and effluent vial 240) is discarded. Because other components of the instrument do not come into contact with the samples, they do not need to be cleaned or disinfected before processing another set of samples with another MOS generation system 200.

[0112] Tracking information, such as an industry-standard 2D barcode (e.g., a QR code or Data Matrix code) or other type of identifier, can be used to track the chain of custody of each patient-specific sample from its origin, to the reservoir 220 containing the sample, and to the effluent vial 240. For example, the instrument 100 can be equipped with barcode scanning capabilities, a Laboratory Information Management System (LIMS), or other tracking technology. In some examples, similar tracking technology is also used to track reagents, such as matrix materials and oils, to ensure, for example, that the proper and unexpired reagents are used.

[0113] 3A and 3B are top and cross-sectional perspective views, respectively, of a microfluidic chip 210 having a droplet generation region 302 and a polymerization region 304. Four independent flow paths are defined through the microfluidic chip 210 for simultaneously processing four biological samples, although other numbers of flow paths may be defined. Each flow path is for processing a different biological sample.

[0114] Each flow path includes an inlet channel 307 defined on a respective inlet finger 308 of the microfluidic chip 210. Each inlet channel 307 includes a port for connection to a corresponding one of the reservoirs 220, which are sources of a first fluid containing patient-specific biomaterial and unpolymerized matrix material. The inlet fingers 308 are spread apart to allow sufficient space for each reservoir (e.g., each outlet finger 324 is separated from an adjacent finger by a gap). Furthermore, the spread apart inlet fingers 308 allow each finger to flex and seal to its corresponding reservoir 220 independently of each other, thereby avoiding tolerance variations that could prevent sealing (e.g., avoiding a situation where a tall reservoir prevents sealing of an adjacent shorter reservoir). Each flow path also includes one or more (here, two) second channels 310, each including a port for connection to an oil source (e.g., oil reservoir 700, see FIG. 7).

[0115] In the droplet generation region 302, a continuous stream of droplets is generated in each channel by merging a first fluid stream containing a biomaterial (e.g., a patient-specific biomaterial) suspended in an unpolymerized matrix material (e.g., a hydrogel) with one or more second fluid streams. The second fluid is a low-viscosity fluid that is immiscible with the first fluid so that an emulsion of the first fluid in the second fluid can be formed. For example, the first fluid can be a hydrophilic (e.g., water-based) fluid, and the second immiscible fluid can be a hydrophobic material such as oil. While this specification may refer specifically to oil or hydrophobic fluids, it should be understood that the disclosed approach generally applies to any suitable second fluid that is immiscible with the first fluid. Generally, the first and second fluid flows are controlled by separate flow regulator devices controlled by respective controllers, such as proportional-integral-derivative (PID) controllers. For example, the flow of the first fluid is controlled by pressurizing the reservoir 220 using, for example, a syringe pump or other suitable flow regulator, and the flow of the second fluid is controlled by a pump having a programmable flow rate.

[0116] When the first and second fluid flows intersect, droplets of the first fluid are formed in the second fluid. Each droplet contains unpolymerized matrix material from the corresponding reservoir and a small amount of biological sample, dispersed (e.g., emulsified) in oil. The droplets have a stable, substantially spherical shape, and the relative flow rates of the first fluid and oil control the size (e.g., diameter, volume) of the droplets. In the polymerization region 304, the matrix material in the droplets is polymerized by exposure to a stimulus, such as heat, light, a chemical stimulus, or another suitable stimulus, thereby forming an emulsion of MOS in oil. After polymerization, the MOS is recovered from the oil by demulsification in a demulsification cartridge 230 (shown in FIG. 2 ) fluidly connected to the microfluidic chip 210. For example, increasing the flow rate of the first fluid relative to the flow rate of the second fluid generates larger droplets. Decreasing the flow rate of the first fluid relative to the flow rate of the second fluid generates smaller droplets. The relative flow rates of the first and second fluids also affect the flow rate of the emulsion containing the droplets produced.

[0117] The microfluidic chip 210 is an injection-molded chip-on-chip device whose top and bottom surfaces (not shown) are sealed with a thin cover, such as a glass cover or a polymer film. The cover is a thin film having a thickness of, for example, 50 μm to 100 μm, e.g., 80 μm. This double-sided geometry allows the flow paths to cross each other, doubling the density of channels in the polymerization region 304 compared to a similarly structured chip with a single-sided geometry. In some examples, for example, if the microfluidic chip 210 is designed for thermal or photoinitiated polymerization, the cover is a film that does not substantially absorb heat or light, respectively. For example, the cover film can be optically transparent (e.g., allowing transmission of at least 50%, at least 60%, at least 80%, at least 90%, or at least 95% of the light intensity) at the wavelength(s) of light used for photoinitiated polymerization (e.g., 405 nm blue light).

[0118] FIG. 4 is an enlarged view of a portion of the droplet generation region included within the dashed-line box in FIG. 3A. As shown in FIG. 4, the first and second channels 306, 310 in each flow path intersect, e.g., at a right angle, at a junction 311 in the droplet generation region 302. As the first fluid exits the first channel 306 and the oil exits the second channel 310 and enters the junction, the convergence of the oil flows (e.g., flowing from both sides of the junction) causes a small amount of the first fluid to break off after passing through the junction, thereby allowing droplets of the first fluid encased in oil to begin forming. From the junction, the droplets in each flow path flow into a respective droplet generation chamber 312. In some examples, the droplet generation chamber 312 is an expanded chamber having a cross-sectional dimension (e.g., cross-sectional area) larger than that of the first channel 306. The transition between the narrow first channel 306 and the wide droplet generation chamber 312 along each flow path slows the fluid flow, increasing pressure and causing droplet breakage. The continuous flow of the first fluid and oil forms a continuous stream of evenly spaced droplets of substantially uniform shape and volume. Each droplet contains patient-specific biomaterial and unpolymerized matrix material.

[0119] From each droplet generation chamber 312, the droplets flow into a narrower outlet channel 314. For example, the height and width of the outlet channel 314 are slightly larger than the droplet diameter, e.g., about 10% to about 25% larger. The reduction in cross-sectional dimension from the droplet generation chamber 312 to the outlet channel 314 creates backpressure to support droplet generation. This reduction in size also forces the droplets to form in a line with substantially uniform spacing, preventing them from contacting each other and coalescing prior to downstream polymerization processes. In one example, the spacing between droplets in the outlet channel 314 can be approximately twice the spacing between droplets in the droplet generation chamber 312.

[0120] In a particular example, for droplets having a diameter of 260 μm, the outlet channel 314 has cross-sectional dimensions of 300 μm×300 μm, the droplet generation chamber 312 has cross-sectional dimensions of 200 μm×700 μm, and the first and second channels 306, 310 have cross-sectional dimensions of 200 μm×200 μm.

[0121] An imaging subsystem (not shown, see FIG. 13 ) is positioned to capture images (e.g., still or video images) of the droplets as they pass through the droplet generation chamber 312, the outlet channel 314, or both. The images can be used to determine droplet characteristics, such as droplet size (e.g., droplet volume or droplet diameter), droplet size distribution, droplet velocity, spacing between adjacent droplets, droplet number density (e.g., number of droplets per unit length of the outlet channel 314, number of droplets per unit volume of the second fluid), an estimated total number of droplets generated from a given starting volume of the first fluid, or other droplet characteristics. In some examples discussed further below, the droplet characteristics can be used for closed-loop feedback control of the droplet generation process. For example, the flow rates of the oil, the first fluid, or both can be adjusted in real time based on the determined droplet characteristics to obtain droplets of a target size.

[0122] In the illustrated example, each outlet channel 314 folds back alongside its corresponding droplet generation chamber 312, e.g., in a serpentine shape, so that a portion of each outlet channel 314 extends in a direction parallel to the corresponding droplet generation chamber 312. This shape allows droplets to remain within the field of view of the imaging subsystem for a longer period of time. Furthermore, droplets can be imaged in a single filed, evenly spaced apart position in the outlet channel 314, facilitating image analysis. Furthermore, the ability to image droplets in adjacent channels of different cross-sectional areas and calculate their velocities enables internal QC, where the velocity ratio corresponds to the channel area ratio. Furthermore, the droplet generation chambers 312 and outlet channels 314 are compressed into a relatively small area of ​​the microfluidic chip 210, which allows the imaging subsystem to operate with a small field of view, thereby capturing images with high resolution, e.g., micron-per-pixel resolution.

[0123] 3A and 3B, droplets flow from each flow path's outlet channel 314 into respective polymerization channels 320 in polymerization region 304. Polymerization channels 320 are configured so that droplets flowing therethrough are exposed to a stimulus, such as heat, light (e.g., blue light), a chemical stimulus, or another suitable stimulus. The stimulus induces polymerization of unpolymerized matrix material in the droplets, thereby forming MOS emulsified in oil. In some examples, the stimulus is applied to both the top and bottom surfaces of microfluidic chip 210; in some examples, the stimulus is applied to only one surface.

[0124] In the illustrated example, the polymerization channel 320 is a serpentine channel, with a portion of the channel length located on the top surface of the microfluidic chip 210 and the remainder located on the bottom surface of the microfluidic chip 210. For example, for a 1-meter long polymerization channel, 0.5 m of the channel length is located on the top surface and 0.5 m of the channel length is located on the bottom surface. This arrangement allows for a high density of polymerization channels 320, thereby providing a long fluid path length and maximizing exposure time to the stimulus. Additionally, the positioning of the polymerization channel 320 on the top and bottom surfaces of the microfluidic chip 210 means that there is only a small separation between the droplets in the channel 320 and the applied stimulus—for example, a separation distance approximately equal to the thickness of the transparent film covering the microfluidic chip. This small separation minimizes delay in the stimulus reaching the polymerization channel, reducing energy loss and contributing to efficient polymerization. The combination of the serpentine path on both surfaces of the microfluidic chip 210 and the small separation distance between the channel 320 and the applied stimulus allows polymerization to be achieved with a relatively short residence time, e.g., 30 seconds to 2 minutes, e.g., 1 minute, along a polymerization channel approximately 1 meter long. Residence time refers to the time it takes for a given droplet to flow along the length of one of the polymerization channels 320.

[0125] The shape of the polymerization channel 320 is designed to contribute to efficient polymerization. For example, the cross-sectional area of ​​the channel 320 may be slightly larger than the cross-sectional area of ​​the droplets. This prevents droplets from stacking on top of each other within the channel and instead keeps them substantially evenly spaced and in a single file, thereby helping to avoid clogging and ensuring that all droplets are evenly exposed to an applied stimulus. The serpentine path of the polymerization channel 320 provides a long path length even within the limited area of ​​the microfluidic chip 210. Furthermore, tight turns, e.g., 180° turns, along the polymerization channel 320 promote mixing of the fluid flow within the channel, thereby facilitating heat transfer by, for example, inducing convection, and promoting polymerization efficiency.

[0126] Parameters for polymerization of the matrix material in the droplets may depend on the size of the droplets. For example, the duration of exposure of the droplets to a stimulus sufficient to induce complete polymerization of the matrix material in the droplets may depend on the size of the droplets. Larger droplets may take longer to polymerize than smaller droplets due to the larger volume of material and the longer distance between the droplet edge and the droplet center. For example, droplets of a particular size may require an exposure time of at least 90 seconds, while droplets of a different, smaller size may require only 30 seconds. The microfluidic system can be adjusted so that the stimulus is sufficient to obtain complete polymerization of the matrix material in droplets of a target size, for example, so that droplets flowing through the system at a target flow rate are exposed to the stimulus for a time sufficient to polymerize droplets of the desired size. Specifically, the length of the polymerization channel 320 is fixed. Therefore, the flow rate (flow velocity) of the droplets along the polymerization channel 320 determines the exposure time of the droplet volume.

[0127] In some examples, the system is designed so that complete polymerization is achieved when droplets of a target size flow through the polymerization channel 320 at a target flow rate. In some examples, various combinations of target sizes and target flow rates are available. For example, the system may be designed so that complete polymerization is achieved when droplets of a first size range flow at a first flow rate, and complete polymerization is achieved when droplets of a second, smaller size range flow at a second, faster flow rate.

[0128] The downstream end of each polymerization channel 320 is fluidly connected to a corresponding outlet channel 322, each of which is connected to a corresponding demulsification cartridge (not shown). Each outlet channel 322 is formed in a respective outlet finger 324 of the microfluidic chip 210. The outlet fingers 324 are spread apart (e.g., each outlet finger 324 is separated from adjacent fingers by a gap) to allow sufficient space for each demulsification cartridge 230. Furthermore, the spread apart outlet fingers 324 allow each finger to flex and seal to the corresponding demulsification cartridge 230 independently of each other finger, thereby avoiding tolerance variations that could interfere with sealing.

[0129] 5A-5C, for thermal polymerization, the matrix material is a temperature-sensitive material, such as Matrigel, that polymerizes in response to exposure to heat. Heat blocks 500a and 500b (collectively referred to as heat block 500) are positioned to apply heat to the top and bottom surfaces of the polymerization region of the microfluidic chip 210. The heat block 500 is designed to expose the droplets in the polymerization channel 320 to a temperature sufficient to induce polymerization but not high enough to destroy the biomaterial, e.g., 35°C to 45°C, e.g., 37°C. Furthermore, the heat block 500 is designed to induce a rapid temperature increase to the target temperature and to provide precise temperature control. In some examples, thermal polymerization of droplets having Matrigel as the polymerizable matrix material can be achieved with an exposure time of less than 2 minutes, e.g., less than 1 minute.

[0130] The heat blocks 500 are secured in place, for example, by clamps. In some examples, the heat blocks 500 are secured in place in a manner that promotes contact with the surface of the microfluidic chip 210. For example, the heat blocks 500 may be spring loaded to press them against the surface of the microfluidic chip 210. In some cases, springs on both sides of the heat blocks 500 are used to secure the heat blocks to the microfluidic chip. For example, a first spring secures the heat block 500a to the microfluidic chip, and a second spring secures the heat block 500b to the microfluidic chip. The heat blocks 500 do not contact the droplet generation region 302 to avoid inducing polymerization of the matrix material before it reaches the polymerization region 304.

[0131] As shown in the cross-sectional view of FIG. 5B and the exploded view of FIG. 5C, each of the heat blocks includes a heating element 502 that contacts a respective surface of the polymerization region 304 of the microfluidic chip 210. The heating elements 502 are surrounded on all sides except the side facing the microfluidic chip 210 by an insulating cover 504, such as a polytetrafluoroethylene (PTFE) or ultra-high molecular weight polyethylene (UHMWPE) cover. Specifically referring to FIG. 5C, in one example, the heating elements 502 are 6 mm thick copper blocks heated on the side opposite the microfluidic chip side by respective resistance heaters 506, such as Kapton film resistance heaters, e.g., Minco 24 VDC resistance heaters. In one example, the resistance heaters 506 are wired in parallel and controlled by a controller under closed-loop feedback control using temperature feedback provided by temperature sensors attached to one or both of the heating elements 502, such as thermistors, thermocouples, resistance temperature detectors, or other suitable temperature sensors. The controller has a serial interface that allows a computer to control the operation of and read the current temperature of the heating elements 502. In some examples, a temperature sensor is attached to each of the heating elements 502, and each heating element 502 is independently controllable by a corresponding thermostat controller, thereby allowing precise temperature control.

[0132] Referring to Figures 6A-6C, when using photo-induced polymerization, the matrix material is a material that polymerizes in response to exposure to light of a specific wavelength, such as 405 nm blue light. For photopolymerization, light blocks 600a, 600b are positioned to illuminate droplets in the polymerization channel of the microfluidic chip. Each light block 600a, 600b defines a respective integration chamber 602a, 602b. The light block 600b houses a light source, such as one or more light-emitting diodes (LEDs) 604, e.g., 405 nm blue LEDs, disposed in the integration chamber 602b. For example, the LEDs 604 are mounted on a printed circuit board (PCB) 606, e.g., a metal-core PCB, to dissipate heat generated by the LEDs. The LEDs can be arranged in a one-dimensional or two-dimensional array. In some examples, LEDs are disposed in both integration chambers 602a, 602b. Heat sinks are placed on the outside of one or both of the optical blocks 600a, 600b to further dissipate heat.

[0133] The integrating chambers 602a, 602b are designed to repeatedly reflect light around and through the microfluidic chip to maximize exposure and exposure uniformity. For example, the integrating chambers 602a, 602b are formed from a material with high diffuse reflectance and no yellowing, such as white PTFE or UHMWPE. The high diffuse reflectance of both integrating chambers 602a, 602b allows light from the LED 604 to pass through the transparent microfluidic chip 210 multiple times. A material with sufficiently high reflectance is one that reflects at least 90%, at least 95%, or at least 99% of visible light and has a consistent reflectance across all visible wavelengths (e.g., a reflectance that varies less than 20%, less than 10%, less than 5%, or less than 1% across the visible wavelength spectrum). Diffuse reflectance refers to the property of a material in which all light rays incident on the material are dispersed in many directions and are not specularly reflected.

[0134] To further facilitate light reflection and prevent light from escaping the integration chambers 602a, 602b, the optical blocks 600a, 600b are fixed in place, e.g., spring-loaded, pressing them against the surface of the microfluidic chip. In some cases, springs on both sides of the optical blocks 600a, 600b are used to secure the optical blocks 600a, 600b relative to the microfluidic chip. For example, a first spring secures the optical block 600a relative to the microfluidic chip, and a second spring secures the optical block 600b relative to the microfluidic chip. A photodetector 608, such as a photodiode, is positioned within the integration chamber 602b of the optical block 600b to measure the intensity of light within the integration chamber 602b. In some examples, the signal from the photodetector is used to verify whether there is sufficient light intensity to complete the light-induced polymerization process; if not, sample processing does not proceed. In some examples, the signal from the photodetector is used for closed-loop computer control of the light source. In some examples, a photodetector is also disposed in the integration chamber 602a.

[0135] In some examples, light-induced polymerization is fast enough that polymerization occurs within less than the residence time of the droplets in the polymerization channel 320. For example, light-induced polymerization can be a photoinitiated process that, once triggered with light of the appropriate wavelength, continues to propagate to completion without continued exposure to light. In these examples, the light source can be pulsed for energy efficiency and to reduce heating of the droplets. The light block is positioned to avoid irradiating the droplet generation region to avoid inducing polymerization of the matrix material before it reaches the polymerization region.

[0136] In some instances, the matrix material is polymerized by chemical polymerization. In chemical polymerization, two liquid materials (e.g., two hydrogels, one of which contains a biomaterial) are mixed, and an emulsion of the mixed hydrogels in a hydrophobic fluid is produced in a droplet generation region. Polymerization begins substantially immediately upon mixing the two hydrogels. The ratio of the biomaterial-containing hydrogel to the other hydrogel is about 1:1 to about 2:1. Droplet size is controlled by the combined flow rate of the two hydrogels.

[0137] In a specific example, a microfluidic chip designed for chemical polymerization contains two independent flow paths for simultaneous processing of two biological samples. Each flow path contains two channels for two hydrogels. The two hydrogel channels merge into a single channel slightly upstream of the junction with the flow path for the hydrophobic fluid. A droplet generation chamber is located downstream of the junction, as shown in Figure 4. The droplet generation chamber and the subsequent outlet channel have a length such that the residence time of the emulsion allows for complete chemical polymerization of the droplets. The emulsion of polymerized droplets in the hydrophobic fluid flows into a demulsification cartridge.

[0138] Referring again to FIGS. 3A-3B, a notch 330 is defined between the droplet generation region 302 and the polymerization region 304 of the microfluidic chip 210. The notch 330 provides thermal and / or optical isolation between these regions to prevent undesired premature polymerization of matrix material still present in the droplet generation region. The notch 330 can have angled sides (see FIG. 2B ), for example, at a 45° angle relative to the surface of the microfluidic chip 210, thereby forming a total internal reflection (TIR) ​​mirror. When light is used as the polymerization stimulus, this TIR mirror reflects the stimulus light away from the droplet generation region 302, thereby protecting the matrix material in the droplet generation region 302 from exposure to the light. When heat is used as the polymerization stimulus, the notch 330 acts as a thermal insulator, protecting the matrix material in the droplet generation region 302 from exposure to the applied heat. For example, the cutout 330 extends through the entire thickness of the microfluidic chip 210 .

[0139] 7A-7C, oil reservoir 700 stores oil used for droplet generation. Oil is poured into chamber 710 of reservoir 700 through an opening at the top of the reservoir, which can be closed by cover 702. In some examples, cover 702 is a pressure cap that is secured in place with toggle clamp 704, allowing the pressure cap to be removed without a twisting motion. Oil exits chamber 710 through port 706 at the bottom of the chamber, e.g., to avoid the use of a dip tube. In some examples, port 706 is a threaded port, facilitating a reliable seal for tubing connected to it.

[0140] Internal retroreflective ribs 708 extend along the height of the chamber 710, allowing for measuring the volume of oil in the chamber 710 in discrete increments, for example. In some examples, the ribs are designed to allow for measuring five different oil volumes. Based on the number and volume of samples being processed, the system can verify that enough oil is present before starting.

[0141] 7D-7E show an example of an oil reservoir 750 for storing oil used in droplet generation. Oil is poured into a chamber 760 of the reservoir through an opening at the top of the reservoir 750, which can be closed by a cover 752, which may be a filter vent cap used to prevent particulates from entering the oil. In some cases, the oil is not pressurized, and the oil reservoir 750 does not include a pressure cap and / or toggle clamp. The oil reservoir 750 is configured with (e.g., includes) an interface 756 that indicates the volume of oil in the oil reservoir 750. The interface 756 may display other measurements instead of or in addition to volume. The oil exits the chamber 760 through a port 758 at the bottom of the chamber 760.

[0142] 8A is a schematic block diagram of an exemplary oil delivery subsystem 800 for delivering oil from reservoir 700 to microfluidic chip 210. Generally, during the droplet generation process, the flow of oil to each of the four droplet generation channels is individually metered, thereby controlling the rate of droplet flow through the polymerization section of the microfluidic chip.

[0143] The oil delivery subsystem 800 includes a pressure regulator 802 that supplies a single fixed air pressure, for example, 2 bar, to the oil reservoir 700, forcing oil out of a port 706. A pressure sensor 804 and an air flow sensor 806, positioned between the pressure regulator 802 and the oil reservoir 700, provide feedback for error checking. A three-way valve 808 is also positioned between the fixed pressure regulator 802 and the oil reservoir 700. The three-way valve 808 has a default open configuration that vents pressure in the oil reservoir 700 and is closed by the controller to pressurize the reservoir 700. This is a safety feature that ensures that the reservoir 700 does not become pressurized when opened for refilling, for example.

[0144] Oil exiting the oil reservoir 700 passes through a manifold 810 equipped with a single check valve 812 to prevent contamination from backflow, then branches into four flow control channels 814. Flow in each of the four channels 814 is controlled by a respective servo flow valve 816 (e.g., an Enidine PFV-W24E01-P050E-0300 servo flow valve) attached to the manifold 810. In some examples, control of the servo valves 816 is via a 0-10 VDC analog control signal, providing closed-loop control of the oil flow rate, for example, to control the size of the droplets generated. Two-way valves 818 are positioned along each channel 814 to function as flow shutoff valves. Downstream of the valves 818, tubing connects the channels 814 to oil ports on the mount for the microfluidic chip 210.

[0145] FIG. 8B is a schematic block diagram of an exemplary oil delivery subsystem 850 for delivering oil from an oil reservoir chamber 760 to the microfluidic chip 210. The oil delivery system includes a manifold 864 for distributing the output from a single oil reservoir to four independent syringe pumps 856, 858, 860, and 862 with programmable flow rates controlled by a controller, such as a PID controller. The output from the four syringe pumps 856, 858, 860, and 862 connects to four oil inlet ports on the droplet generation chip. Each syringe has sufficient capacity to perform the entire droplet generation without stopping and refilling. The oil delivery subsystem 850 includes one or more level sensors 854 configured to measure the volume of oil in the oil chamber 760 and an oil cap sensor 852. In one specific example, the oil reservoir 750 includes five level sensors 854.

[0146] 9A-9B, each biological sample is packaged in a closed, disposable (single-use) reservoir 220. The reservoir 220 is designed for transport from the sample preparation area to the MOS generation instrument and for direct insertion into the instrument. The reservoir 220 includes a plastic base 902 and a plastic top cover 903, e.g., a polypropylene base and top cover. The base 902 and top cover 903 are attached together to form an airtight seal, e.g., by ultrasonic welding. The base 902 defines a single well 910 within the reservoir 220, having a capacity for a single biological sample, e.g., less than 5 mL, less than 2 mL, less than 1 mL, or less than 500 μL, and more than 5 μL, more than 10 μL, or more than 20 μL. The well has a sloping bottom 906 that slopes from the bottom 906 of the well 910 to a lowest point below a tube 912 that extends to the cover 904. This shape facilitates processing of a high percentage of the sample, e.g., substantially 100%, contained in well 910. This shape also facilitates loading reservoir 220 into a MOS generating instrument in the proper orientation.

[0147] The reservoir 220 includes a flexible cover 904, e.g., an elastomeric cover, disposed over a top cover 903. An inlet port 920 and an outlet port 922 are defined in the flexible cover 904. The inlet port 920 and the outlet port 922 are flexible seals that are fluidly sealed by the flexible cover 904 when not in use. The inlet port 920 is, e.g., a duckbill valve. To insert a biological sample into the well 910 of the reservoir 220, the operator pushes the pipette tip through a slit in the duckbill valve of the inlet port 920 to inject the sample. Air gaps along the slit on either side of the pipette tip allow air to escape as the sample is injected. When the pipette is withdrawn, the duckbill valve closes and seals to prevent leakage.

[0148] Additionally, the duckbill valve of the entry port 920 is sized and shaped to mate with a pressure manifold when the reservoir is positioned in the MOS generating device. This configuration allows the well 910 to be pressurized without penetrating the duckbill valve of the entry port, thereby reducing the risk of contamination of the biological sample contained in the well 910.

[0149] Outlet port 922 is a seal that connects the reservoir outlet to the microfluidic chip inlet. When well 910 is pressurized, the sample contained in well 910 is forced through tubing 912 and out through outlet port 922, from where the sample enters microfluidic chip 210.

[0150] 10A and 10B, samples from multiple reservoirs 220 (e.g., up to four reservoirs in the illustrated example) can be processed simultaneously by the MOS generation instrument. Each reservoir 220 is fluidly connected via its outlet port 922 (FIGS. 8A-8B) to the inlet port of one of the fluid flow paths through the microfluidic chip 210. In some examples, a set of multiple reservoirs 220 is handled together in a carrier 150, for example, for transport to the MOS generation instrument. The carrier 150 can have an identifier 152, such as a barcode (e.g., a QR code or data matrix), serial number, or other identifier printed or affixed thereto, that can be used to identify the samples contained in the reservoirs 220 carried by the carrier 150. In some examples, each reservoir 220 also has an identifier that can be used to identify the sample contained therein.

[0151] Carrier 150 resides in sample preparation station 151 for sample preparation prior to transport to the MOS generation instrument. Sample preparation station 151 houses sample tubes 153 containing biological material (e.g., minced tissue) suspended in processing medium or buffer, and media tubes 155 containing fresh medium for dilution. The user dilutes the sample in each sample tube 153 to the desired concentration using the medium in the corresponding media tube 155 and then transfers the diluted suspension to the corresponding reservoir 220.

[0152] In some examples, it is important to maintain the biological sample at a sufficiently low temperature before processing in the MOS-generating instrument. During transport to the instrument, the carrier 150 carrying the reservoir 220 can be cooled, for example, by placing the carrier in an ice bath or by using another cooling mechanism. With reference to FIGS. 11A and 11B, in some examples, the MOS-generating instrument includes a cold block 154 that receives the carrier 150 carrying the reservoir 220 and keeps the biological sample cool while it is in the instrument. In one example, the cold block 154 is cooled by a thermoelectric cooling subsystem 156 that includes a thermoelectric cooler 158 with a heat sink 160 (e.g., a pin-fin heat sink) and a fan 162 to dissipate heat.

[0153] 12, the sample drive subsystem 170 applies air pressure independently to each reservoir 220 to drive the biological sample from the reservoir through tubing 912 (FIG. 9A) and into the corresponding channel of the microfluidic chip 210. The air pressure applied to each reservoir can be controlled independently from the air pressure applied to each of the other reservoirs, thereby independently controlling the fluid flow rate along each channel. Because the fluid flow rate affects the size of droplets generated in the microfluidic chip, in some examples, the sample drive subsystem 170 performs closed-loop feedback control of the fluid flow rate to achieve a target droplet size, for example, based on droplet size measurements obtained from the imaging subsystem.

[0154] The compressed air manifold 172 serves as a central distribution and monitoring point for supplying air to multiple subsystems, such as pressure servo valves 174. Incorporating a pressure sensor 173 within the manifold allows the system to prevent operation if the supply pressure is too low. The pressure servo valves 174 are, for example, Elveflow regulators with a maximum value of 2 bar and a resolution of 0.0001 bar. The pressure servo valves 174 are controlled by a controller 176, such as a proportional-integral-derivative (PID) controller. An air flow sensor 178 and a two-way valve 180 are positioned in series between each servo valve 174 and its respective reservoir 220. The air flow sensor 178 provides feedback on the air flow rate, for example, for error detection. The two-way valves 180 act as shutoff valves to prevent backflow when the pressure regulator is shut off.

[0155] In the illustrated example, the sample drive subsystem 170 is shown implemented using a pressure servo valve. In some examples, other types of flow regulators, such as pressure regulators, valves, or pumps, such as peristaltic pumps, diaphragm pumps, syringe pumps, or other suitable flow regulators, are used. Control of the flow regulators allows precise variation of the pressurization in each reservoir, for example, in increments of 0.1 mbar or 0.01 mbar.

[0156] Generally, the sample drive subsystem 170 is operable to control fluid flow rates in the range of approximately 100-200 μL / min by pressurizing the reservoir using air pressures of approximately 50-800 mbar. The sample drive subsystem 170 allows continuous fluid flow for the total sample volume, for example, sample volumes of approximately 10 μL-1 mL. The sample drive subsystem 170 can implement a purge function involving higher air pressures, for example, air pressures up to approximately 2000 mbar.

[0157] 13A and 13B, the imaging subsystem 350 is positioned to capture images (e.g., still or video images) of droplets as they pass through the droplet generation chamber 312 and outlet channel 314 (see FIGS. 3A-3B) of the microfluidic chip 210. The images captured by the imaging subsystem are used for measurement or other characterization of the droplets and / or for user visualization of the generated droplets as part of a real-time closed-loop flow control feedback system. For example, droplet characteristics determined based on the images can be provided as input to a closed-loop control system that controls the flow rates of the first fluid, the second fluid, or both to achieve droplets with target characteristics (e.g., a target size) or to obtain a desired number of droplets (e.g., for a screening library).

[0158] The imaging subsystem 350 includes a camera 352, such as a still camera or video camera, a lens 354 attached to the camera 352, a mirror 356 on a kinematic mount 358, and a backlight 360. The camera 352, lens 354, and mirror 356 on the mount 358 are positioned on one side of the microfluidic chip 210. The backlight 360 is positioned on the other side of the microfluidic chip 210. A mount (not shown) houses the camera 352, lens 354, and mirror 356. In some examples, the imaging subsystem 350 is separated from the microfluidic chip by a transparent window, e.g., a glass window, to protect the components of the imaging subsystem from droplets and debris. Images captured by the camera 352 are provided to a computing device, such as a local computer or a cloud-based server, having one or more processors coupled to memory.

[0159] Generally, camera 352 and lens 354 have a field of view of approximately 20 mm x 25 mm, a pixel resolution of approximately 5 μm, and a frame rate of 42 frames per second. Imaging subsystem 350 can be a monochromatic or polychromatic imaging system. Imaging subsystem 350 has a global shutter, exhibits no parallax or distortion in the outer channels, and is capable of sub-frame region-of-interest transfer.

[0160] The backlight 360 is positioned to illuminate the droplet generation chamber and outlet channel of the microfluidic chip 210 so that images can be captured. The backlight 360 can be monochromatic or polychromatic and illuminates the chip 210 with a wavelength that does not induce polymerization and does not heat the droplets sufficiently to induce polymerization. In a particular example, the light is a green light mounted at least 75 mm from the chip. In some examples, multiple light sources of different colors are used to facilitate image analysis. The liquids, including the first fluid containing the patient-specific biomaterial and unpolymerized matrix material and the second immiscible fluid, are generally both transparent liquids, and the droplets are visible in the immiscible fluid due to the refractive index difference between the droplets and the surrounding fluid. The curvature of the droplet edges, combined with the refractive index difference, causes the droplets to act as lenses that bend the light from the backlight 360.

[0161] In some examples, the backlight 360 is a diffuse light source, which makes the edges of the droplets visible in the image captured by the camera 352. In some examples, the light is a collimated light source to facilitate image analysis. For example, when illuminated with collimated light, the focusing effect of the lenticular droplets deflects the light from its original axis and out of the line of sight of the camera 352. This focusing effect increases the contrast of the droplets in the resulting image compared to droplets illuminated with diffuse light, which may facilitate identification of the droplets and / or their edges in the image. In some examples, for example, in place of the backlight 360, a light is positioned on the same side of the microfluidic chip 210 as the camera.

[0162] In some examples, the imaging subsystem uses strobe illumination to capture multiple exposures of each of one or more droplets within a single image captured during a single frame of camera 352. The timing of the two exposures, e.g., the spacing between the two exposures and the duration of each exposure, ensures that the same droplet is captured in both exposures without artifacts such as blurring that may result from subject motion within the image. The two exposures of the same droplet are analyzed to determine droplet characteristics, such as droplet size (e.g., droplet volume or droplet diameter), droplet size distribution, droplet velocity, spacing between adjacent droplets, droplet number density (e.g., number of droplets per unit length of the outlet channel (e.g., outlet channel 314 in FIG. 4 ), number of droplets per unit volume of the second fluid), an estimated total number of droplets generated from a given starting volume of the first fluid, or other droplet characteristics. These properties are applied to a closed-loop feedback system that can adjust the flow rate of the feedstock (e.g., the first fluid or the second fluid) to affect the size and flow rate of the droplets produced, thereby enabling the continuous production of droplets of a target size at a target flow rate, as described in more detail below.

[0163] After polymerization, the polymerized droplets pass through a demulsification subsystem, which transfers the droplets from an emulsion of droplets in a second fluid (e.g., oil) to an aqueous fluid in a continuous, microfluidic-based demulsification process. The demulsification subsystem is designed to minimize droplet loss, e.g., achieve at least 90% droplet recovery into the aqueous fluid, and transfer the droplets into the aqueous fluid without physical damage to the droplets. The demulsification subsystem is also designed to prevent residual oil from being transferred into the aqueous fluid. Furthermore, the demulsification subsystem can achieve high-throughput processing, e.g., processing input volumes ranging from 10 to 1,000 μL, and has the ability to remove large volumes of oil, e.g., up to 5 mL of oil.

[0164] FIG. 14 is a schematic diagram of a portion of a demulsification subsystem 450. The demulsification subsystem 450 uses cross-flow filtration in a microfluidic chip format, where a combined stream of aqueous fluid and oil containing polymerized droplets enters the inlet end 451 of a microfluidic channel 452 and flows along the channel 452 above a membrane 454. The membrane is a hydrophobic and lipophilic membrane, such as a polyvinylidene fluoride (PVDF) or PTFE membrane. A pressure bias across the membrane 454 drives the oil through the membrane 454 and into a collection reservoir 456 below the membrane. For example, a vacuum can be applied to the surface of the membrane 454 facing the collection reservoir 456. In some examples, the collection reservoir 456 is vented to atmosphere. In some examples, a positive pressure is applied to the surface of the membrane 454 facing the microfluidic channel 452. In some examples, the oil is driven through the membrane by the pressure difference between the positive pressure applied above the membrane and the ambient pressure below the membrane, aided by gravity.

[0165] Because membrane 454 is hydrophobic, the aqueous fluid and polymerized droplets are repelled from membrane 454 and therefore continue to flow along microfluidic channel 452 to the outlet end 455 of the microfluidic channel.

[0166] As the oil is drawn through the membrane, the polymerized droplets are displaced into the aqueous fluid. By the outlet end 455 of the microfluidic channel, the oil has been drawn through the membrane so that only polymerized droplets in the aqueous fluid remain in the channel.

[0167] The microfluidic channel 452 of the demulsification subsystem 450 is serpentine to promote mixing of the oil and aqueous fluid within the channel, thereby facilitating contact between the oil and the membrane 454 even when the oil occupies a small percentage of the fluid volume (e.g., toward the outlet end 455 of the microfluidic channel).

[0168] FIG. 15 shows a cross-sectional view of a portion of a demulsification subsystem 450. A microfluidic channel 452 is defined in a substrate 470, such as a molded plastic substrate (e.g., polymethyl methacrylate (PMMA) or polystyrene). A first side of a hydrophobic membrane 454 (e.g., a PVDF membrane) is attached to the substrate 470 with an adhesive 472, which is cut out to expose the microfluidic channel 452 in the substrate 470. The opposite side of the membrane 454 is attached to a collection reservoir 456, such as a plastic, e.g., PMMA, reservoir, with an adhesive 474. The adhesive 474 is positioned around the outer edge of the membrane 454 to allow oil in the microfluidic channel to be drawn through the membrane 454 and into the collection reservoir 456. The collection reservoir 456 itself has a central chamber 453 located below the portion of the membrane not covered by the adhesive 474, so that oil drawn through the membrane 454 is collected within the cavity of the collection reservoir 456. The adhesives 472, 474 are biocompatible adhesives that are inert to oil and water-based fluids.

[0169] Membrane 454 has pores that are sized to prevent the passage of droplets while readily allowing oil flow through the membrane, for example, the pores have a diameter of 0.25 μm to about 1 μm, e.g., 0.45 μm.

[0170] 16A-16C are top, side, and exploded views, respectively, of an exemplary demulsification cartridge 230. A sample inlet channel 460, a media inlet channel 462, and a microfluidic channel 452 are defined in a substrate 470. The substrate 470 is connected to a collection reservoir 456, with a membrane 454 disposed therebetween, such that the membrane 454 forms the bottom wall of the microfluidic channel 452. A media reservoir 476 is disposed below the collection reservoir 456. The substrate, collection reservoir 456, and media reservoir 476 are connected, for example, by ultrasonic welding.

[0171] The microfluidic channel 452 is defined on a surface of the substrate 470 facing the collection reservoir 456 such that fluid in the microfluidic channel 452 contacts the membrane 454 and such that oil is drawn through the microfluidic channel 452 into the collection reservoir 456. In some examples, the media inlet channel 462 is defined on the same surface of the substrate 470 as the microfluidic channel 452, and in some examples, the media inlet channel 462 is defined on the opposite surface of the substrate 470.

[0172] The collection reservoir 456, e.g., a molded plastic (e.g., PMMA or polystyrene) structure, defines one or more chambers 453 for holding oil removed from fluid flowing along the microfluidic channel 452. Generally, the volume of this chamber within the collection reservoir 456 is larger than the expected volume of oil to be removed from the fluid, e.g., about 10-25% larger. For example, if the expected volume of oil is about 5 mL, the chamber volume may be about 6 mL. In some examples, the collection reservoir 456 includes an oil-trapping material, such as a sponge-like material, disposed on the bottom surface of the chamber. The oil-trapping material is a hydrophobic material that traps oil within the chamber 453 of the collection reservoir 456, thereby preventing the oil from being drawn back through the membrane 454 into the microfluidic channel 452. In some examples, a vacuum channel is defined in the body of the collection reservoir 456 to apply a vacuum to the membrane 454.

[0173] The media reservoir 476, e.g., a molded plastic (e.g., PMMA or polystyrene) structure, defines a chamber that holds the aqueous fluid delivered to the microfluidic channel 452. In some examples, this chamber of the media reservoir 476 has a capacity of approximately 10-20 mL, e.g., 4-8 mL, of aqueous fluid. In some cases, the capacity of the media reservoir 476 is limited by the volume of the final effluent tubing. For example, if the capacity of the final effluent tubing is 15 mL, the volume of the media reservoir chamber should not exceed 15 mL. The bottom surface 478 of the chamber is angled relative to the plane of the substrate 470 so that the aqueous fluid accumulates below the sipper tube 480. The aqueous fluid is drawn from the media reservoir 476 via the sipper tube 480, which is fluidly connected to the media inlet channel 462. In the illustrated example, the sipper tube 480 is integral with the collection vessel 456 and sealed to the substrate 470 with an adhesive. In some examples, the sipper tube is a separate element that extends through a hole in the collection container 456. In some examples, the sipper tube 480 is integral with the substrate 470 (e.g., molded as part of the substrate 470) so as to avoid the use of a separate sealing element between the sipper tube 480 and the substrate 470.

[0174] The media reservoirs 476 can be filled with aqueous fluid through valves 484, which extend through openings 486, 488 in the substrate 470 and collection container 456, respectively. In some examples, the valves 484 are molded duckbill valves, e.g., thermoplastic elastomer valves, which allow a pipette tip to be inserted through the valves 484 to fill the media reservoirs 476 but which otherwise remain closed to prevent leakage or contamination of the aqueous fluid. The valves 484 also form a seal against the pressure port and can open to allow the entry of air pressure to drive the flow of aqueous fluid into the media inlet channels 462. A media pressure subsystem, such as a set of precision pressure regulators protected by shut-off valves, provides individual driving pressures to each media reservoir 476, thereby driving the flow of aqueous fluid into the media inlet channels 462.

[0175] The seal 490 is positioned to provide a sealed connection between the microfluidic chip 210 and the sample inlet channel 460. The seal 490 may be a thermoplastic elastomer seal. In the illustrated example, the valve 484 and the seal 490 are a single, integral element. In some examples, the valve 484 and the seal 490 are two separate elements.

[0176] As previously discussed, the sample inlet channel 460 is fluidly connected to the outlet of the polymerization section (see FIGS. 3A-3B) and the microfluidic channel 452. The media inlet channel 462, which is configured to receive an aqueous fluid through a valve, is also connected to the microfluidic channel 452 such that the fluid flowing along the microfluidic channel 452 near the inlet end 451 is a mixture of aqueous fluid and droplets in oil. In some examples, the sample inlet channel 460 and the media inlet channel 462 have a three-dimensional channel structure that is elevated above the bottom surface of the substrate 470, for example, thus preventing the oil and aqueous fluids from contacting the membrane 454 until they meet at the microfluidic channel 452. In some examples, the media inlet channel 462 is a long channel (e.g., approximately 50 mm to 150 mm in length) with small cross-sectional dimensions (e.g., 125 μm × 225 μm), allowing operation at high backpressures, thereby preventing oil from backflowing into the microfluidic channel 452 when the flow over the membrane 454 slows.

[0177] The first section 492 of the microfluidic channel 452 has a simple serpentine pattern that allows the initial, oil-laden fluid to wet the membrane 454, thereby facilitating rapid removal of oil from the fluid flowing within the channel 452. The second section 494 of the microfluidic channel 452 has a double serpentine pattern with small folds (e.g., folds 495) formed within each round trip of the channel 452. These small folds facilitate mixing of the fluid within the channel such that a shrinking bolus of oil in the fluid repeatedly contacts the membrane 454. This repeated contact helps prevent situations in which a small oil bolus (e.g., an oil bolus less than a certain percentage of the channel depth, e.g., less than 50%) fails to contact the membrane 454 along the entire length of the channel 452.

[0178] In some examples, the cross-sectional dimensions (e.g., cross-sectional area) of the microfluidic channel 452 decrease along the length of the channel, increasing fluid resistance and thereby increasing fluid residence time on the membrane 454 and facilitating complete removal of the oil. This tapering also helps maintain backpressure against the membrane 454 as the fluid volume decreases as the oil is removed. In a particular example, the microfluidic channel 452 tapers from initial dimensions of 300 μm × 600 μm to final dimensions of 300 μm × 325 μm. The tapering can be gradual or in discrete increments.

[0179] At the outlet end 455 of the microfluidic channel 452, droplets in the substantially oil-free aqueous fluid flow through an outlet channel 496 to an outlet port 498 extending, for example, through a shield element 499 integrally formed with the collection container 456. The shield element 499 prevents contamination of the edge of an effluent container connected to the outlet port 498, which would otherwise occur due to splashing caused by process air pushing the liquid out of the outlet port 498. The outlet channel 496 can be a long channel, for example, approximately 50-100 mm long, with cross-sectional dimensions similar to the final cross-sectional dimensions of the microfluidic channel 452, creating a backpressure equivalent to that of the external tubing of a breadboard. In some examples, the outlet port 498 has a pipette-like tip configured to connect to an effluent container, such as a conical effluent container (see FIG. 18). The demulsified MOS can be transported in the effluent container to another location, for example, for further organoid growth and testing.

[0180] 17A-17B show two exemplary geometries of a serpentine microfluidic channel 452, including a single serpentine channel 452a (17A) and a double serpentine channel 452b (17B). Other geometries are possible. In each geometry, an emulsion of droplets in oil is received via a sample inlet channel 460, and an aqueous fluid is received via one or more media inlet channels 462. At the outlet end 455 of the microfluidic channel 452, the droplets in the aqueous fluid exit via an outlet port 498 into an effluent vial, such as a centrifuge tube (not shown).

[0181] Components of the MOS generating instrument that come into contact with the sample (e.g., patient-specific biomaterial) are disposable. These components, and other disposable components, are referred to as consumables. With reference to Figures 18A and 18B, the consumables are positioned, for example, in a holder or nest, so that the various fluid interfaces are aligned and fluidly connected.

[0182] The positioned consumable parts are clamped in place by clamp 850, the interface seals are pressed together, and the air supply is mated with the air actuation port. For example, clamp 850 is lowered by the action of an actuator such as a lever. In the example of FIGS. 18A and 18B, clamp 850 includes a spring plunger that presses certain components together to provide a good seal. For example, pressure supply 854 and chip entry port 856 press down on the top of each sample reservoir 220. Oil port 852 presses down on microfluidic chip 210. Pressure supply 858 presses down on demulsifier cartridge 230 to connect it to media reservoir 476. Plunger 860 applies a downward force to the demulsifier cartridge to keep it in the nest when sealing pressure is applied on the opposite side.

[0183] 19, in some examples, loading of consumables follows a predetermined sequence. The sample and media reservoirs of the demulsification cartridge are pre-filled off-instrument (50). Prior to operation, a sample identifier is associated with each location of the sample reservoir carrier (52). Effluent vials are placed in an effluent vial holder, and the effluent vial identifier (e.g., barcode) is read (e.g., scanned) as each vial is loaded, thereby linking each effluent vial identifier to a corresponding channel (54).

[0184] The demulsifying cartridges are loaded into a demulsifying cartridge holder, and the outlet port of each demulsifying cartridge is inserted into a corresponding effluent vial 56. The sample reservoir carrier, which contains multiple sample reservoirs, is read (e.g., scanned) and placed in the instrument's cooled inlet nest 58, thereby linking the sample identifier of each sample with the corresponding channel and, therefore, with the corresponding specific effluent vial identifier.

[0185] The microfluidic chip is placed into the instrument and aligned and mated with the sample reservoir and demulsification cartridge (60). The consumable parts are clamped together to form seals against the open ports of the various components (62).

[0186] Referring to FIG. 20 , droplet generation system 200 operates under the control of an electronic subsystem 250 implemented on a computing device (e.g., local or cloud-based) that communicates with various electronically controlled subsystems via an interface, such as a USB or I2C serial interface. The electronic subsystem provides a graphical user interface that enables development, diagnostic testing, and service of the droplet generation system. The electronic subsystem is also operable to control the hardware of droplet generation system 200 to process samples according to a selected protocol and communicate results, for example, to a graphical user interface or another computing device. In some examples, the electronic subsystem performs closed-loop feedback control of the operation of droplet generation system 200 based on analysis of data received from the droplet generation system. For example, the electronic subsystem can determine the size of droplets generated based on images acquired by the imaging subsystem and control the droplet generation system to adjust the fluid flow rate through droplet generation chamber 212 to achieve a target droplet size.

[0187] The electronic subsystem 250 includes user interface components 252, such as a display 254 (e.g., a touchscreen display), a keyboard 256, a mouse 258, and a barcode reader 259. Additional and / or alternative user interface components may also be included. The electronic subsystem 250 includes data acquisition components, including data input and data output components. Digital output components include controls for functions such as two-way valves (e.g., for solenoid valves or syringe pumps), heat, and light. Digital input components include, for example, a door closure sensor. Analog input components include, for example, a polymerization light intensity sensor and a leak sensor.

[0188] The electronic subsystem 250 includes devices that connect to the controlling computing device through one or more types of interfaces. In the example of Figure 21, these devices include USB devices 270 such as a camera, a pressure regulator (e.g., an Elveflow precision pressure regulator), a temperature controller, a strobe controller, and a barcode reader, and I2C sensor input devices 272 such as pressure, airflow, and temperature sensors.

[0189] Additional components for the electronic subsystem may include, for example, a power entry module, a power supply, a USB hub, an Ethernet bulkhead pass-through, a cooling fan, and other components, as shown in FIG.

[0190] FIG. 21 is a block diagram of a MOS generation system showing the incorporation of certain elements of the electronic subsystem.

[0191] 22, in an exemplary method for MOS generation (e.g., generation and demulsification), droplets of a first fluid (e.g., a patient-specific biomaterial in a polymerizable matrix material) are generated in a hydrophobic fluid, such as oil, in a droplet generation portion of a first microfluidic channel defined on a surface of a microfluidic chip (10). For example, the droplets are generated at a junction between the first microfluidic channel and one or more channels carrying the hydrophobic fluid. In some examples, the microfluidic chip has multiple fluidically independent first microfluidic channels, and droplets are generated in parallel (e.g., simultaneously) in each first microfluidic channel, e.g., to enable simultaneous processing of samples from multiple patients.

[0192] In some examples, to obtain droplets of a desired size, the flow rate of the first fluid (a fluid containing the biomaterial suspended in the unpolymerized matrix material) is controlled (12), e.g., by a closed-loop feedback control system. For example, the flow rate is controlled based on the size of the generated droplets, determined from images of the droplets in the droplet generation chamber of the first microfluidic channel. The flow rate of the hydrophobic fluid is also controlled (13) to control the rate of the emulsion flow (e.g., the flow of fluid through the outlet channel 314; see Figure 4), which determines the fluid residence time in the polymerization region (e.g., polymerization channel 320; see Figures 3A-3B). The flow rate of the emulsion flow is also related to the adjustment of the start time and rate of medium introduction into the demulsification channel (e.g., via the medium inlet channel 462; see Figures 16A-16C).

[0193] The generated droplets flow through a polymerization portion of a first microfluidic channel (14), where a stimulus, e.g., light or heat, is applied to polymerize the matrix material, thereby forming MOS emulsified in the hydrophobic fluid (16). When multiple fluidically independent first microfluidic channels are present, droplets in each channel polymerize simultaneously as they flow through the polymerization portion of their respective first microfluidic channels. In some examples, when light is applied as the stimulus, the droplets are illuminated with a pulsed illumination pattern. In some examples, the stimulus (e.g., light or heat) is applied to two opposing sides of the microfluidic chip.

[0194] The mixture of the aqueous fluid and the emulsion of MOS in the hydrophobic fluid is flowed along a second microfluidic channel defined in the substrate of the demulsification cartridge (18) for demulsification of the MOS, e.g., to transfer the MOS into a suspension in the aqueous fluid. When there are multiple first microfluidic channels defined in the microfluidic chip, each first microfluidic channel is fluidly connected to a respective second microfluidic channel in the corresponding demulsification cartridge. As the mixture flows along the second microfluidic channel, the hydrophobic fluid is transferred through a membrane forming the wall of the second microfluidic channel, thereby removing the hydrophobic fluid and allowing the MOS to be transferred to the aqueous fluid (20). In some examples, a vacuum is applied to the side of the membrane opposite the second microfluidic channel to facilitate removal of the hydrophobic fluid from the second microfluidic channel.

[0195] The suspension of MOS in aqueous fluid is provided to an effluent vial (22) for downstream applications, such as cell growth, testing, or other uses.

[0196] Closed-loop control of MOS generation systems.

[0197] In some examples, real-time closed-loop feedback is used for real-time quantification and control of droplet size and flow rate in the above-described MOS generation systems and methods. Based on the measured droplet size and flow rate, the microfluidic system can be controlled to adjust the size of the droplets generated, the droplet flow rate, or both. These adjustments can achieve a target droplet size, flow rate, or both to enable efficient and complete polymerization of the MOS matrix material. These adjustments can also control the droplet generation process to generate a target number of droplets, for example, to obtain a desired number of MOSs for a screening library.

[0198] These approaches to closed-loop feedback may have one or more of the following advantages: The closed-loop feedback systems described herein provide a mechanism for reliably and noninvasively capturing multiple images of a single droplet as it flows along a microfluidic channel. Quantitative measurements of the size and velocity of a single droplet, as well as an estimate of the total number of droplets, can be obtained using short exposure times and high-intensity illumination pulses. Images are acquired without artifacts such as blurring due to droplet motion. The timing of the illumination pulses is controllable independently of the camera shutter and is independent of the camera frame rate, allowing it to be adjusted to achieve consistent image quality over a wide range of flow rates. The ability to sense physical parameters such as size and velocity with a non-contact, non-invasive approach increases the accuracy of processes occurring in microfluidic devices. Additionally, these imaging approaches can be achieved using low camera frame rates, resulting in a relatively low computational load for analyzing the generated images. Furthermore, the optical system configurations used in the closed-loop feedback approaches described herein are less expensive and smaller than typical optical systems used in life science systems because, for example, the optical systems described herein do not require expensive features such as submillimeter alignment, photochromic dyes, dichroic components, or lasers. Additionally, the closed-loop feedback systems described herein are non-invasive to microfluidic systems, thereby avoiding the possibility of clogging or contamination.

[0199] In the closed-loop feedback approach described herein, multiple exposures of one or more droplets are acquired within a single image, and the droplet size, droplet flow rate, or both are determined based on the exposures. The timing of the two exposures, e.g., the spacing between the two exposures and the duration of each exposure, ensures that the same droplet is captured in both exposures without artifacts such as blurring that may result from subject movement within the image. The two exposures of the same droplet are analyzed to determine droplet characteristics, such as size and flow rate. The exposure images may also be analyzed to determine other droplet characteristics, such as the spacing between adjacent droplets, the droplet number density, or an estimated total number of droplets generated by the system. These characteristics are applied to a closed-loop feedback system that can adjust the first and second fluids accordingly to generate droplets of a target size, to cause the generated droplets to flow at a target flow rate, or both, thereby enabling the continuous generation of droplets of a target size at a target flow rate. In a particular example, the flow rate of the first fluid is adjusted to achieve droplets of a target size, and the flow rate of the second fluid is adjusted so that the generated droplets flow at the target flow rate.

[0200] In some examples, a closed-loop feedback approach can be used to control the number of droplets generated in a microfluidic system. For example, a microfluidic system can have a performance goal that indicates a target number of droplets of a particular size to be generated from a given volume of sample, such as generating 10,000 droplets of a specified size from a 10 μL sample of biological material from a needle biopsy. Based on the droplet sizes determined from multiple images, the flow rates of the first and second fluids can be adjusted to achieve the target number of droplets.

[0201] 13A and 13B , the closed-loop feedback system includes an imaging subsystem 350, a computing device, such as a local computer or cloud-based server having one or more processors coupled to a memory, a controller controlling the flow of the first fluid and the second fluid, and a controller controlling the operation of the imaging system 350. The closed-loop feedback system captures and analyzes images of droplets to quantify the size of the generated droplets, the flow rate of the generated droplets, or both. In some examples, the closed-loop feedback system controls operating parameters of the MOS generation system based on analysis of the images to obtain droplets of a target size, a target flow rate, or both, suitable for complete and efficient polymerization. In some examples, the closed-loop feedback system controls operating parameters of the MOS generation system based on analysis of the images to obtain a target number of droplets from a given sample of the first fluid. For example, the closed-loop feedback system controls the operation of flow regulators that control the flow of the first and second fluids, e.g., controls the operation of the sample drive subsystem 170 (see FIG. 12) to control the flow of the first fluid, and / or controls the operation of one or more of the programmable flow pumps 856-862 (see FIG. 8B) to control the flow of the second fluid.

[0202] The liquids, including the first fluid containing the patient-specific biomaterial and unpolymerized matrix material, and the second immiscible fluid, are generally both clear liquids, and the droplets are visible in the immiscible fluid due to the refractive index difference between the droplets and the surrounding fluid. The curvature of the droplet edges, combined with the refractive index difference, causes the droplets to act as lenses that bend light from light source 360. In some examples, light source 360 ​​is a diffuse light source, which makes the droplet edges visible in images captured by camera 352. In some examples, light source 360 ​​is a collimated light source, and the focusing effect of the lenticular droplets deflects light from its original axis and away from the line of sight of camera 352. This focusing effect increases the contrast of the droplets compared to droplets illuminated with diffuse light.

[0203] The operation of the camera 352 and light source 360 ​​is controlled by a controller in a closed-loop feedback system, such as a programmable control board capable of generating multiple simultaneous outputs of control pulses (e.g., current or voltage pulses) for synchronized control of the camera and light source with programmable timing.

[0204] The camera 352 includes a shutter, e.g., a global shutter. For example, the shutter can be an electronic shutter internal to the camera. The opening of the shutter can be controlled by an external trigger, such as a current or voltage pulse received from a controller in a closed-loop feedback system. The exposure time (e.g., the amount of time the shutter remains open) can be specified, for example, by direct interaction with the camera 352 or by interaction with a computing device communicatively coupled to the camera 352. In a particular example, the shutter of the camera 352 responds to a transistor-transistor logic (TTL) input, e.g., a 3.3V or 5V TTL input, received from a controller. When the shutter receives the input from the controller, the shutter opens and an integration time lasting the specified exposure time begins. A single integration time, such as a single still photographic frame or a single frame of a video camera, may be referred to as a single frame of the camera 352.

[0205] Light source 360 ​​can be controlled by an external trigger, such as a current or voltage pulse received from a controller of a closed-loop feedback system. Light source 360 ​​is a light source capable of stroboscopic illumination, e.g., generating multiple (e.g., two, three, four, or more) short illumination pulses in rapid succession. For example, light source 360 ​​can be a light-emitting diode (LED), an arc lamp, or another suitable light source. In some examples, light source 360 ​​is a collimated light source, e.g., a collimated LED. Light source 360 ​​can be a monochromatic light source (e.g., a monochromatic LED) or a polychromatic light source (e.g., a white light source). In certain examples, a monochromatic blue LED is used as the light source. The short wavelength of the blue LED helps prevent blurring of the edges of the droplets in the image. In some examples, such as when the droplets contain an absorbing dye, light source 360 ​​includes multiple light sources of different colors, e.g., multiple different colored LEDs.

[0206] The controller of the closed-loop feedback system is configured to control the camera 352 and the light source 360 ​​so that the light source 360 ​​generates multiple (e.g., two, three, four, or more) illumination pulses during a single frame of the camera 352, e.g., while the camera shutter remains open. This results in a double exposure within a frame: two exposure images of the droplet generation chamber 312 (see FIGS. 3A-3B ) (within which there may be droplets) within a single image captured during a single frame of the camera 352. In some examples, the exit channel 314 is also within the field of view of the camera 352 and is therefore also captured by the two exposures. The two exposures are separated in time by the time between two successive illumination pulses generated by the light source 360. Due to the short time interval between the two illumination pulses, the same droplet(s) appear in both exposure images. A given droplet appears in slightly different positions in the two exposures due to the flow of droplets along the droplet generation chamber 312 during the time interval between the two illumination pulses.

[0207] In some examples, the camera 352 operates at a fixed frame rate and outputs a TTL signal at the beginning of each frame integration (e.g., when the shutter opens). The controller of the closed-loop feedback system is programmed to generate strobe output pulses with a specific start delay, duration, and interval. In some examples, the controller of the closed-loop feedback system triggers the camera exposure and strobe pulses so that the time between exposures can be varied. In some examples, a modified sequence can be implemented in which the strobe is fired in an alternating sequence of a single pulse in one frame followed by a double pulse in the next frame. This modified sequence allows the system to collect two image streams, one containing double-exposure images for analysis, for example, for speed, and the other containing single-exposure images for display to a user, for example.

[0208] The ability to image droplets in both the droplet generation chamber 312 and the exit channel 314 has advantages. Specifically, because the shape of the droplet generation chamber 312 differs from the shape of the exit channel 314, the ability to image droplets in both locations provides the opportunity to image droplets having different shapes. In one example, the first and second channels 306, 310 have cross-sectional dimensions of 200 μm by 200 μm, the droplet generation chamber 312 has cross-sectional dimensions of 200 μm by 700 μm, and the exit channel 314 has cross-sectional dimensions of 300 μm by 300 μm. If the nominal spherical droplet size is 255 μm, the droplets will flatten into a pancake-shaped shape with circular edges while in the droplet generation chamber 312 and become spherical upon exiting the droplet generation chamber 312 and entering the exit channel 314. The ability to image droplets in both the droplet generation chamber 312 and the exit channel 314 allows the droplet to be imaged in both shapes, facilitating volume calculations, for example.

[0209] Additionally, the ability to image droplets and determine droplet velocity in both the production chamber and the exit channel provides a mechanism for error checking. Because the production chamber and the exit channel have different cross-sectional areas, the expected velocity ratio between the two locations is known (e.g., the velocity ratio is the inverse ratio of the cross-sectional areas). Comparing the measured droplet velocities at the two locations can act as an error check, for example, to prevent a single-location measurement from reporting an erroneous velocity and instead facilitate reporting a velocity consistent with data from both channels.

[0210] FIG. 23A shows an example of a single image 1200 captured during one frame of the imaging system's camera 352 (FIGS. 13A-13B). Image 1200 includes an image of a portion of a production chamber with two captured locations 1202a, 1202b of the same droplet. The two locations 1202a, 1202b of the droplet were captured as a double-exposure image by two successive illumination pulses while the camera shutter remained open. The two locations 1202a, 1202b in a single image may be referred to as two exposure images 1202a, 1202b of the same droplet. Because the droplet is flowing along the production chamber in the direction indicated by the arrows, the droplet appears in a different location in each of the two exposure images 1202a, 1202b. The position of the droplet in exposure image 1202a is the position of the droplet when the first flash of illumination is generated, and the position of the droplet in exposure image 1202b is the position of the droplet when the second flash of illumination is generated. Because the droplet has moved along the chamber in the interval between flashes, it is in a different position in each of the two exposure images 1202a, 1202b. As discussed further below, analysis of multiple exposure images 1202a, 1202b of a single droplet in a frame can be performed to determine droplet characteristics such as droplet size, droplet velocity, or other properties.

[0211] FIG. 23B shows an example of control signals provided from a controller to the camera and light source that result in the generation of the double exposure image of FIG. 23A. For example, a controller in a closed-loop feedback system can generate the control signals of FIG. 23B to control the operation of the shutter of camera 352 (see FIGS. 13A-13B) and the operation of light source 360. The control signals include camera drive signals 1210, e.g., square wave current pulses, that cause the shutter of camera 352 to open. The control signals also include light source drive signals 1220, e.g., square wave current pulses, where each drive signal 1220 causes light source 360 ​​to generate an illumination pulse. The center line in FIG. 23B indicates the camera integration time 1230, e.g., the time that the shutter of camera 352 remains open.

[0212] In the example of Figures 23A and 23B, the controller first sends a camera drive signal 1212 to control the camera shutter to open, thereby initiating a single camera frame (integration time) for capturing a single image 1200. Next, the controller sends two successive light source drive signals 1222a and 1222b, each of which causes the light source to generate a short flash of illumination. The camera shutter remains open during both illumination pulses, as indicated by the camera integration time 1230. Each flash of illumination generates one exposure image of the droplet within the single image 1200 (e.g., the flashes triggered by pulses 1222a and 1222b result in exposure images 1202a and 1202b of the droplet, respectively). After the integration time 230 ends, the camera shutter closes. The resulting image 1200, including the two exposure images 1202a and 1202b, is sent to a computing device for analysis.

[0213] In some examples, the time interval between the two pulses 1222a, 1222b is short enough that the distance between the leading edge 1204a of a droplet in the exposure image 1202a and the leading edge 1204b of the same droplet in the subsequent exposure image 1202b is less than the size (e.g., radius or diameter) of the droplet. This short time interval ensures that the same droplet appears in both exposure images 1202a, 1202b (e.g., a given droplet travels a distance less than its diameter during the time interval between pulses). For example, the time interval between the two pulses 1222a, 1222b (pulse interval) can be 1 millisecond (ms) to 50 ms, e.g., 1 ms to 30 ms, or 5 ms to 10 ms. The time interval can be adjusted by programming the controller to accommodate various flow rates, for example. For example, the time interval can be adjusted to provide a consistent distance between the leading edges of droplets in each image, regardless of the flow rate. The time interval may be consistent throughout the imaging process or may vary.

[0214] In some examples, the time interval between the two pulses 1222a, 1222b has a duration such that the droplet moves from about one-quarter of its radius to about four times its radius between exposures. Droplet identification and analysis can be achieved even if the droplet does not overlap itself between the two exposures. This flexibility allows image analysis over a wide dynamic range of speeds.

[0215] The pulses 1222a, 1222b can be short enough to avoid artifacts (e.g., blurring) due to droplet movement in the respective exposed images 1202a, 1202b. The duration of each pulse can be set based on system factors such as lens focusing, lens aperture setting, camera image sensor sensitivity, and camera gain. For example, each pulse 1222a, 1222b can have a duration of about 5 microseconds (μsec) to about 125 μsec, e.g., about 10 μsec to about 50 μsec, or about 25 μsec to about 50 μsec, e.g., 10 μsec, 20 μsec, 25 μsec, 30 μsec, 35 μsec, 40 μsec, 45 μsec, 50 μsec, 75 μsec, 100 μsec, or 120 μsec. In a specific example, the two pulses 1222a, 1222b are 125 μsec pulses separated by a 25 ms time interval. The pulse durations (pulse intervals) do not necessarily have the same value. Because the pulses 1222a, 1222b are short, high intensity light can be used, which makes image analysis easier.

[0216] The multiple exposures 1202 a, 1202 b in a given image 1200 are analyzed by a computing device having one or more processors coupled to memory to perform image analysis, for example, using machine vision analysis processing techniques. The analysis can determine characteristics of the droplet(s) in the exposures 1202 a, 1202 b, such as droplet size (diameter, volume), droplet flow rate (velocity), spacing between adjacent droplets, droplet number density, or an estimated total number of droplets generated from a particular volume of sample.

[0217] The computer analysis may include identifying droplets in each exposure 1202a, 1202b of a given image 1200 and identifying the leading or trailing edge of the droplets in each exposure (e.g., leading edge 1204a of the droplets in exposure 1202a and leading edge 1204b of the droplets in exposure 1202b). A leading edge of a droplet is the edge of the droplet facing the direction of droplet movement, and a trailing edge is the edge facing opposite the direction of droplet movement. In some examples, the leading edge, trailing edge, or both edges of the droplets in each exposure are directly identified without first identifying the droplets themselves.

[0218] In one example, droplets are identified in the exposed image by a frequency domain technique or machine vision analysis process that identifies circular or substantially circular objects (two-dimensional projections of spherical droplets) in the exposed image and, for example, creates a best-fit circle. Frequency domain techniques, such as autocorrelation or fast Fourier transform processes, may allow for a holistic analysis of part or all of each image. In some cases, the machine vision analysis process identifies circular or substantially circular features within a pre-specified target size range, for example, to avoid falsely identifying foreign objects such as debris or bubbles as droplets.

[0219] In one example, the leading edge of a droplet is identified in the exposed image as a feature having a positive curvature in a pre-specified direction corresponding to the direction of droplet movement, and the trailing edge of the droplet is identified as a feature having a negative curvature in the same direction. In some cases, the machine vision process identifies features having a curvature within a pre-specified range of preferred curvatures, for example, to avoid falsely identifying foreign objects as droplets.

[0220] In some examples, after identifying droplets in each exposure image, the exposure images 1202a, 1202b in the image 1200 are analyzed to determine the distance between corresponding edges of the droplets in two consecutive exposure images 1202a, 1202b. In the example of Figure 23A, the distance d between the leading edge 1204a of the droplet in exposure image 1202a and the leading edge 1204b of the droplet in exposure image 1202b is determined. In some examples, the distance between the trailing edges of the droplets in two consecutive exposure images 1202a, 1202b is determined. The distance between corresponding edges of the same droplet in two consecutive exposure images is the distance the droplet traveled along the imaging channel during the time between two consecutive illumination pulses triggered by the two light source drive signals 1222a, 1222b. Since the time interval between the two light source drive signals 1222a, 1222b is known, the velocity of the droplet can be determined as the distance traveled by the droplet divided by the time interval between the two light source drive signals 1222a, 1222b. The velocity of a given droplet is the flow rate of the droplet along the imaging channel.

[0221] In some examples, after identifying droplets in each exposure, exposures 1202a, 1202b of image 1200 are analyzed to determine droplet size. In the example of FIG. 23A , droplet diameter D is determined as the distance between leading edge 1204a of the droplet in exposure 1202a and trailing edge 1206a of the droplet in the same exposure 1202a. Additionally, droplet volume can be determined based on the determined droplet diameter, e.g., based on the expected or observed shape of the droplet given the shape of the channel. An estimate of the total number of droplets generated from a given starting volume of sample can be determined based on the determined volume of a single droplet.

[0222] The number density of droplets per unit length of the channel or unit volume of the second fluid can be determined based on the distance between adjacent droplets, for example, the spacing between the trailing edge of a first droplet and the leading edge of the next successive droplet in the channel.

[0223] The closed-loop feedback system controls the operating parameters of the MOS generation system 200 based on the droplet characteristics determined from the image analysis. Specifically, a computing device communicates with a controller controlling the flow of the first fluid, the second fluid, or both to control the flow rates of the first fluid, the second fluid, or both, for example, to adjust the size of droplets generated at the junction 311 (see FIG. 4 ), the flow rate of droplets through the droplet generation chamber 312, or both. In a particular example, the flow rate of the first fluid is adjusted to achieve droplets of a target size, and the flow rate of the second fluid is adjusted so that the generated droplets flow at a target flow rate. For example, if the image analysis reveals that droplets larger than a target size are generated, the computing device communicates with the controller controlling the flow rate of the first fluid to decrease the flow rate of the first fluid. If the image analysis reveals that the flow rate of droplets along the droplet generation chamber 312 is greater than the target flow rate, the computing device communicates with the controller controlling the flow rate of the second fluid to decrease the flow rate of the second fluid, or both.

[0224] In one embodiment, the controllers controlling the flow rates of the first and second fluids are implemented as PID controllers to facilitate stable control. This implementation allows for interactions between system parameters to be considered, for example, providing for independent control of oil flow rate and sample pressure despite the dependency between these two parameters. For example, one PID controller considers velocity and oil flow rate, while another PID controller considers droplet size and sample pressure, thus allowing for stable control of both droplet size and velocity.

[0225] The camera in the closed-loop feedback system can be operated at a frame rate of 1 Hz to 30 Hz. In some examples, the camera frame rate is faster than the cycle rate of the flows of the first and second fluids. In this way, dynamic adjustments can be made in real time to maintain the generation of droplets of a target size at a target flow rate. In a specific example, the flows of the first and second fluids are controlled by pressure with a cycle rate of 2 Hz to 5 Hz. In this example, a camera with a frame rate of 15 Hz, 30 Hz, or 40 Hz can be used to capture multiple images. For example, if multiple image streams (e.g., a double-exposure image stream and a single-exposure image stream) are collected at a frame rate of 30 to 40 Hz, the double-exposure image stream for analysis can be collected at 15 to 20 Hz, and a single-exposure image stream for display to the user, interleaved with the double-exposure image stream, can also be collected at 15 to 20 Hz.

[0226] In some examples, a closed-loop feedback system can determine when a performance problem occurs in the MOS generation system 200. For example, if the supply of the first fluid is depleted, droplets will not form and therefore will not be identified in the image of the droplet generation chamber 312. To avoid circulating air bubbles from an empty reservoir of the first fluid (e.g., reservoir 220; see FIGS. 3A-3B), the closed-loop feedback system can control the flow regulator of the fluid flow from reservoir 220 to shut off the flow from reservoir 220 while maintaining the flow of the second fluid from oil reservoir 700 at a target flow rate. For example, a two-way shut-off valve can be positioned between the flow regulator for the reservoir of the first fluid (e.g., a syringe pump) and reservoir 220 to prevent pressure from the second fluid from causing backflow into reservoir 220. The continuous flow of the second fluid ensures that already generated droplets flow the entire length of the outlet channel 314 at the target flow rate for complete polymerization.

[0227] 24A, in some examples, the light source 1354 of the imaging system includes multiple individual light sources 1356a, 1356b, each emitting a different color of light. For example, the light source 1356a may be a first LED that emits red light, and the light source 1356b may be a second LED that emits green light.

[0228] FIG. 24B shows the drive signals of an imaging system including the multiple individual light sources 1356a, 1356b of FIG. 24A. The camera drive signal 1210 and integration time 1230 of the camera are as described above with respect to FIG. 23B. In this example, a controller sends light source drive signals 1320 to the light sources. The light source drive signal 1320 includes a first light source drive signal 1322a sent to a first individual light source 1356a, causing the first light source to emit light of that color (e.g., red light). The light source drive signal 1320 also includes a second light source drive signal 1322b sent to a second individual light source 1356b, causing the second light source to emit light of that color (e.g., green light). In this manner, two successive illumination pulses of two different colors are generated.

[0229] The two exposures produced by successive illumination pulses of two different colors within a single frame are one color (e.g., red) and the second color (e.g., green). This color difference facilitates identification of the droplets in each exposure: the red droplets are from the first exposure and the green droplets are from the second exposure.

[0230] 25, in an exemplary process for generating a MOS, a first fluid is flowed through a first microfluidic channel of a MOS generation system (1400). The first fluid includes a biomaterial, e.g., a patient-derived biomaterial, and an unpolymerized matrix material. A second fluid is flowed through one or more second microfluidic channels of a microfluidic device (1402). The first and second fluids are immiscible.

[0231] The first fluid and the second fluid are combined at a junction where the first microfluidic channel and the second microfluidic channel meet, thereby forming droplets of the first fluid dispersed in the second fluid 1404. The droplets of the first fluid in the second fluid flow from the junction into a third microfluidic channel.

[0232] A shutter of an imaging device is controlled to open 1406. While the shutter remains open, the light source illuminates 1408 a region of the third microfluidic channel with multiple successive illumination pulses such that multiple exposure images of a single droplet of the first fluid in the third microfluidic channel are captured 1410 in a single image captured by the imaging device.

[0233] The multiple exposure images of a single droplet are analyzed to identify the droplet in each image, or to identify the leading edge, trailing edge, or both of the droplet in each exposure image (1412). In some examples, the image analysis is performed using frequency domain techniques, machine vision techniques, or by creating a best fit circle. The multiple exposure images of a single droplet are analyzed to determine droplet characteristics, such as droplet size, droplet velocity, spacing between adjacent droplets, droplet number density, an estimated total number of droplets generated by the system, or the droplet generation rate (e.g., droplets per second) (1414).

[0234] Based on the determined characteristics of the droplets, the flow rate of the first fluid in the first microfluidic channel, the flow rate of the second fluid in one or more second microfluidic channels, or both are controlled 1416. For example, the flow rate of the first fluid can be controlled to obtain droplets of a target size, and the flow rate of the second fluid can be controlled to obtain droplets flowing at a target velocity.

[0235] The droplets are exposed to a stimulus sufficient to polymerize the matrix material in the droplets (1418), and the polymerized droplets are used (1420), for example, to assay a patient-specific therapy.

[0236] Droplet size measurement and control

[0237] Images of droplets generated in conjunction with the above-described MOS generation systems and methods can be processed to identify any imaged droplets and estimate the size of the imaged droplet(s). Based on the estimated size of the imaged droplet(s), a control system is then configured to adjust the flow rate of at least one fluid in the device (e.g., by adjusting the pressure of the flow of at least one fluid in the device) to dynamically control the size of the droplets subsequently generated. These approaches to droplet size measurement and control can be used in combination with aspects of the above-described MOS generation systems and methods, including in combination with aspects of the above-described closed-loop feedback approach.

[0238] Various embodiments of the approaches to droplet size measurement and control described herein may provide one or more of the following advantages.

[0239] By implementing a feedback system based on the size of the droplets generated, a robust microfluidic system can be provided that adapts to various operating conditions. For example, the techniques described herein can ensure that droplets within a target size range are consistently generated despite variations in the purity of the introduced material, the viscosity of the introduced material, the temperature of the device, etc.

[0240] The techniques described herein may also have the advantage of more quickly identifying the appropriate pressure to apply to a fluid inlet reservoir (sometimes referred to herein as a "holding chamber") to affect fluid flow rate within the device and generate droplets of a target size. Some devices rely on a careful and time-consuming experimental process to determine the appropriate pressure for a very specific set of operating conditions. Unlike such devices, the techniques described herein enable rapid convergence (e.g., within a matter of seconds) on the appropriate pressure for a specific set of operating conditions through real-time, automated pressure adjustment. For simplicity, this specification describes exemplary embodiments of the invention in which fluid flow rate is adjusted by changing the pressure applied to a fluid inlet reservoir. However, it should be understood that the techniques of the present disclosure are equally applicable to other suitable approaches for adjusting fluid flow rate, including, for example, adjusting a programmable pump flow rate applied to the fluid.

[0241] Another advantage of the techniques described herein is that they enable sub-pixel radial resolution for circle detection, which is used to identify and estimate the size of imaged droplets. The term “circle detection” is used herein to refer to identifying circular or near-circular shapes in an image (e.g., based on detected edges in the image) and fitting a circle to represent such a shape (e.g., using a polynomial representation). For example, even if a droplet does not appear exactly circular in an image, the circle detection process can detect the near-circular shape of a droplet and generate a corresponding circular representation for the droplet. In some cases, shapes other than circles can be used to represent near-circular shapes. For example, in some embodiments, polygons with 10 sides, 50 sides, 100 sides, 500 sides, etc. can be used to represent near-circular shapes. In some embodiments, in addition to or as an alternative to polynomial representations, circular or near-circular shapes can be represented by their spectral decomposition, their angular symmetry, etc. Compared to other shape detection algorithms that implement a single-pass approach, the techniques described herein use a multi-pass approach that results in a circle representation with sub-pixel radial resolution. This advantage is particularly important for microfluidic applications, where the imaged droplets (sometimes simply referred to herein as "droplets") may in some cases have radii in the range of only 5-20 pixels (although in other instances the radii may be up to 50 pixels, up to 100 pixels, up to 250 pixels, etc.).

[0242] In some embodiments, the techniques described herein provide various filtering criteria that can potentially avoid undesirable results when processing actual images of droplets in a microfluidic device. For example, filtering criteria implemented by the techniques disclosed herein can prevent false positive detections of circles in spaces between droplets, prevent false positive detections of intersecting circles, filter out detections of circles located too close to the walls of the droplet generating device, and filter out detections of circles with anomalous signals inside their perimeters.

[0243] Furthermore, in some embodiments, the techniques disclosed herein can be used to identify not only droplets but also air bubbles generated by the device. Detection of the air bubbles can indicate a fault condition or near depletion of the introduced material and can be used to control the flow rate of at least one fluid within the device accordingly (e.g., by controlling the pressure applied to a fluid reservoir, which affects the fluid flow rate). For example, in response to identifying an air bubble within the device, the pressure of the fluid flow can be reduced to avoid continuous air flow into the channels of the device.

[0244] In general, in some applications, it may be desirable to control the size of droplets generated by a device. For example, when generating MOS, it may be desirable to ensure that the MOS is generated to a specific size (e.g., 220 microns to 300 microns). If the MOS is too small, there may not be enough space for cells to grow and divide, resulting in a lack of nutrient medium within the MOS. Also, MOS tend to shrink over time, and growing cell clusters may burst droplets that are too small. This bursting may prevent tracking individual cells within the MOS and identify individual cellular responses. On the other hand, if the MOS is generated too large, there may be an excess of nutrient medium within the MOS, resulting in a lack of cells, which may slow the diffusion of drugs to the cells. Furthermore, excessively large MOS may clog the channels of a device (e.g., a microfluidic chip), which may have a channel width of approximately 300 microns in some locations.

[0245] To achieve a target droplet size, existing devices for droplet generation require specific conditions to operate reliably. As described in more detail below, droplet size is sensitive to the flow rates of one or more fluid streams within the droplet generation device (e.g., the ratio of the flow rates of two fluid streams). Therefore, it is desirable to control the flow rates, which can be affected by various factors, including, for example, the viscosity of the fluid, the surface tension of the fluid, the pressure applied to the fluid, and one or more dimensions of the fluid flow path. In some implementations, the flow rates can be controlled directly using a positive displacement pump (e.g., a syringe pump). In other implementations, such as the examples described herein, the flow rate of each fluid can be controlled by adjusting the pressure applied to a corresponding input fluid reservoir that feeds the fluid flow path within the droplet generation device. Thus, while this application describes examples in which fluid flow rates are controlled by applying pressure to an input fluid reservoir, the techniques described herein are equally applicable to other flow rate control mechanisms.

[0246] In examples where the flow rate is controlled by adjusting the pressure applied to the input fluid reservoirs, the appropriate pressure applied to each reservoir can be highly dependent on various operating conditions. For example, in existing devices, consistent purity of the input material, consistent viscosity of the input material, precise temperature conditions, and the like may be required to ensure a consistent desired droplet size. The appropriate pressure to apply under a very specific set of conditions is often determined through a time-consuming experimental process, which can make the droplet generation process vulnerable to slight changes in operating conditions. Therefore, consistent and robust generation of droplets of a target size is a difficult technical challenge.

[0247] 26A shows an example of a device 2100 for forming droplets (e.g., MOS). Device 2100 can be, for example, an implementation of the aspect of MOS generation system 200 discussed above. Device 2100 can include an inlet for introducing either an unpolymerized mixture (already combined) of cells and fluid matrix material, or can receive cells and fluid matrix material separately (e.g., in a holding solution). As noted above, some droplet applications may not include cells at all.

[0248] In some variations, device 2100 includes a holding chamber 2106 for holding the unpolymerized mixture, and / or a holding chamber (not shown) for holding a cell sample (e.g., tumor cells) and for holding a fluid matrix material. For example, holding chamber 2106 may be implemented as reservoir 220 discussed above. Any or all of these holding chambers may be pressurized to control and / or accelerate the flow of fluid out of the chambers and into the device. The device may receive the unpolymerized mixture or may receive and mix components. In some variations, the device can control the concentration of cells in the unpolymerized mixture and can dilute the mixture (e.g., to achieve a desired density by adding additional fluid matrix material. For example, device 2100 can include one or more sensors 2128 for reading the density (e.g., optical density) of the cells in the unpolymerized mixture. The sensors can also be coupled to controller 2124, which can automatically or semi-automatically (e.g., by indicating to a user) control the dilution of the cells in the unpolymerized mixture. Device 2100 can also include a port for receiving the unpolymerized mixture. The port can include a valve or can be coupled to a valve, which can be controlled by controller 2124 (or a separate controller).

[0249] Device 2100 may include another holding chamber 2108 and / or port for holding and / or receiving an immiscible or hydrophobic fluid, such as oil. For example, holding chamber 2108 may be implemented as oil reservoir 700 described above. In some variations, the immiscible fluid may be held in a pressurized chamber such that the flow rate of the immiscible fluid within device 2100 may be controlled. Any of the pressurized chambers may be controlled by controller 2124, which can use one or more pumps 2126 to control the pressure and therefore the flow through device 2100. One or more pressure and / or flow sensors (e.g., sensor 2128) may be included in the system to monitor the flow through the device.

[0250] 26A, the entire device 2100 may be contained within a housing, or a portion of the device 2104 may be contained within a housing. In some variations, the housing may include one or more openings or accesses on the device 2100, for example, for adding immiscible fluids and / or unpolymerized mixtures to the chambers 2106, 2108.

[0251] As mentioned above, device 2100 may also include one or more sensors 2128 for monitoring all or critical portions of the droplet generation process. In some variations, the sensors may include optical sensors, mechanical sensors, voltage and / or resistance (or capacitance, or inductance) sensors, force sensors, temperature sensors, air mass flow sensors, liquid mass flow sensors, pressure sensors, etc. These sensors may be used to monitor the ongoing operation of the assembly, including the formation of droplets. For example, device 2100 may include optical sensor 2132 among sensors 2128. Optical sensor 2132 may be a camera configured to capture images (e.g., photographs or videos) of droplets generated by device 2100, which can be used to control the size of droplets generated by device 2100, as described below. For example, optical sensor 2132 may be part of the closed-loop feedback system discussed above. Device 2100 may also include one or more heat / temperature regulators 2118 for controlling the temperature of either or both of the immiscible fluid and / or the unpolymerized mixture (and / or the fluid matrix material). In some embodiments, monitoring sensor 2128 and / or optical sensor 2132 may be external to device 2100.

[0252] Device 2100 may also include one or more droplet formation assemblies 2120, which may be monitored (e.g., using one or more sensors 2128), as described in more detail below. The droplet formation assemblies may include or be coupled to a dispenser (e.g., a droplet dispenser) 2122. The dispenser may dispense, for example, into one or more collection tubes or multiwell plates 2116.

[0253] Generally, droplet forming assembly 2120 can include one or more microfluidic chips (e.g., microfluidic chip 2130 shown in FIG. 26B or microfluidic chip 210 of FIG. 2) or structures that form and control the flow of the unpolymerized mixture and form the actual droplets. FIG. 26B shows one example of a microfluidic chip 2130 for forming droplets; other examples of microfluidic chips are shown and discussed above, for example, in connection with microfluidic chip 210. In FIG. 26B, chip 2130 includes a pair of parallel structures for forming droplets (e.g., MOS). In other embodiments, the microfluidic chip can include additional structures (e.g., three structures, four structures, five structures, etc.).

[0254] 26C shows a droplet formation region of microfluidic chip 2130 for forming a MOS, which includes a pre-polymerized channel outlet 2141 that opens into outlet channel 2139 and is intersected by immiscible fluid outlet(s) or channel 2143, 2143′. In some embodiments, outlet channel 2139 can be wider than pre-polymerized channel outlet 2141, creating flow slowdown and backpressure to aid droplet formation. In some embodiments, junction region 2137 can be configured as a “+” junction, where immiscible fluid channels 2143, 2143′ intersect with pre-polymerized channel outlet 2141 and outlet channel 2139 to form four right angles. In some embodiments, intersection region 2137 can be configured such that a single immiscible fluid channel (e.g., immiscible fluid channel 2143) enters a straight sample channel (e.g., pre-polymerized channel outlet 2141) at a right angle. In some variations, the input from the immiscible fluid channel(s) can be configured to intersect the unpolymerized material at a non-perpendicular angle. In Figure 26C, as with all figures herein that show dimensions, the dimensions shown are merely illustrative and are not intended to be limiting unless expressly stated otherwise. Other embodiments of droplet formation regions are discussed above in connection with microfluidic chip 210.

[0255] In Figure 26B, microfluidic chip 2130 includes an inlet (introduction port) 2133 for an immiscible fluid into the chip (e.g., from the inlet port or storage chamber shown in Figure 26A). A second inlet port 2135 into the chip can be configured to receive unpolymerized material and transport it to the junction region in a semi-serpentine path. Similarly, the inlet port for the immiscible fluid can be securely coupled to the immiscible fluid chamber or inlet-to-outlet described above.

[0256] Inlet port 2135 to chip 2130 for unpolymerized material may be coupled via a delivery path connecting inlet 2135 to junction region 2137 (as shown in FIG. 26C). Similarly, inlet 2133 for immiscible fluid may connect to two (or more) connecting channels 2143, 2143′ to junction region 2137. The channels leaving junction region 2137 may pass formed droplets (in the immiscible fluid) down the channels to outlet 2131, which may connect to a dispenser (e.g., dispenser 2122 shown in FIG. 26A) for dispensing the droplets into one or more chambers (e.g., multiwell plate 2116 shown in FIG. 26A) for incubation and / or assay.

[0257] In the example shown in Figures 26B and 26C, the formed droplets (e.g., unpolymerized MOS) can be sent downstream through a long, temperature-controlled microfluidic environment (e.g., channel 2139) before being dispensed from the device.

[0258] 26B and 26C, junction region 2137 is shaped as described above so that channel 2141 carrying the unpolymerized mixture intersects one or more (e.g., two) channels 2143, 2143′ carrying a hydrophobic fluid (e.g., oil) that is immiscible with the unpolymerized mixture. When the unpolymerized mixture is pressurized to flow from first channel 2141 at a first velocity, the immiscible fluid flowing in the intersecting channels 2143, 2143′ allows a predetermined amount of the unpolymerized mixture to pass through, after which the unpolymerized mixture breaks off to form droplets that are sent to outlet channel 2139. The size of the droplets is therefore dependent on, and highly sensitive to, changes in pressure (and corresponding flow rates) of both the unpolymerized mixture and the immiscible fluid.

[0259] In some variations, a minced (e.g., dissociated) clinical (e.g., biopsy or resection) sample of dissociated cells, e.g., <1 mm in diameter, can be mixed with a temperature-sensitive gel (i.e., MATRIGEL at 4°C) to form an unpolymerized mixture. This unpolymerized mixture can be placed in a device 100 that generates droplets of a specific volume and material composition, and dissociated cells (e.g., tumor cells) can be dispensed into these droplets. In some cases, each droplet can be approximately 260 microns in diameter (e.g., 220 microns to 300 microns) and contain a tumor-like mass containing 2 to 100 tumor cells (e.g., 2 to 15 tumor cells, 15 to 50 tumor cells, 50 to 100 tumor cells, etc.). The gel in the unpolymerized material can solidify upon heating (e.g., at 37°C) to form a polymerized MOS. In some variations, this method can be used to generate more than 10,000 (e.g., more than 20,000, more than 30,000, more than 40,000, more than 50,000, more than 60,000, more than 70,000, more than 80,000, more than 90,000, more than 100,000) MOSs from a single biopsy. These MOSs are compatible with traditional 3D cell culture techniques and can be used to observe cellular responses to various stimuli via assays, as described above.

[0260] 27A-27E are example images 2200A-2200E of a droplet generation device (such as device 2100 of FIGS. 26A-26C) in operation. Images 2200A-2200E each show outlet channel 2239, which provides an outlet from junction region 2237. The depicted outlet channel 2239 and junction region 2237 may correspond to outlet channel 2139 and junction region 2137, respectively, of microfluidic chip 2130 described in connection with FIGS. 26B and 26C. For example, outlet channel 2239 and junction region 2237 may correspond to droplet generation chamber 312 and junction 311, respectively, of FIG. 4. The images were captured by a camera (corresponding to optical sensor 2132 of device 2100) to monitor the droplet generation process of the device. Each of images 2200A-2200E includes an overlay representing the output of an image processing module that may be implemented by a controller of a droplet generation device (e.g., controller 124 of device 100 shown in FIG. 1A). The image processing module and the steps it performs are described below with reference to FIGS. 28, 429, and 32.

[0261] 27A-27E, overlay 2250 corresponds to detected walls of an imaged exit channel 2239 (sometimes referred to herein simply as "exit channel 2239") in an expanded region (e.g., chamber) of the exit channel 2239. Overlay 2252 corresponds to edges detected within the expanded region by an image processing module (e.g., an edge detection algorithm such as Canny edge detection, Gaussian edge detection, threshold-based edge detection, etc.). Overlay 2254 corresponds to circles detected by the image processing module based on the detected edges (e.g., using a Hough transform).

[0262] In some implementations, it may be beneficial to capture an image of the splayed region and process the content of the image within the splayed region because droplets generated by the device may be less likely to interact with the sidewalls of the outlet channel if they are in this region. Therefore, it may be useful to capture images of generated droplets within the splayed region to estimate the size of the droplets in the absence of interaction with the sidewalls of the outlet channel 2239.

[0263] In some embodiments, it may be useful to capture images of generated droplets in other portions of the system, such as the constricted portion 2260 of the outlet channel 2239. Capturing images of the constricted portion 2260 of the outlet channel 2239 may be beneficial in some cases because droplets in this portion are more likely to be aligned and separated from one another (e.g., by an immiscible fluid). In some embodiments, the constricted portion 2260 may also be deeper than the widened region of the outlet channel 2239 (e.g., 300 microns deep compared to 200 microns deep), allowing droplet size to be measured without compression by the upper or lower walls of the outlet channel 2239. In some embodiments, the constricted portion 2260 may be wide enough so that droplets within the constricted portion 2260 are not compressed by the side walls of the constricted portion 2260.

[0264] In Figures 27A-27B, overlay 2252 shows the presence of several droplets within the expanded region of outlet channel 2239. However, only two circles are detected in each of images 2200A and 2200B (circles 2202 and 2204 in image 2200A, and circles 2206 and 2208 in image 2200B). In some embodiments, this behavior may be desirable because false positives may be much less desirable than false negatives for estimating the size of generated droplets. Because consecutively generated droplets are often likely to have similar sizes and because droplets are produced rapidly (e.g., 40 droplets per second in some embodiments), detecting only a few representative droplets (e.g., fewer than five) within outlet channel 2239 may be sufficient to monitor and control droplet size. Typically, detecting circles that do not actually correspond to droplets may be undesirable because the sizes of these false positives may bias estimates of the true droplet size within outlet channel 2239.

[0265] 27A and 27B, it is also readily observable that circles 2202, 2204 are larger than circles 2206, 2208 in image 2200B, demonstrating the image processing module's ability to accurately detect circles (and corresponding droplets) of various sizes.

[0266] Similar to FIGS. 27A and 27B, FIG. 27C shows image 2200C with detected edges (overlay 2254) indicating the presence of various spherical objects within the expanded area of ​​outlet channel 2239. In FIG. 27C, the image processing module detected three circles 2210 based on these detected edges. However, unlike FIGS. 27A-27B, the detected circles 2210 in image 2200C represent gas bubbles rather than liquid droplets. The imaged gas bubbles (e.g., circle 2210) can be distinguished from the liquid droplets (e.g., detected circles 2202, 2204, 2206, and 2208) based on the darkness of the imaged edges. For example, the edge in image 2200C corresponding to detected circle 2210 is much darker than the edges in images 2200A and 2200B corresponding to detected circles 2202, 2204, 2206, and 2208. This is because the difference in refractive index between air and an immiscible fluid (e.g., oil) is greater than the difference in refractive index between the unpolymerized mixture and the immiscible fluid. Thus, in an exemplary implementation, imaged bubbles can be distinguished from liquid droplets by estimating, for each detected circle, a value reflecting the intensity of the detected circle's periphery (e.g., the ratio of the average intensity on the periphery to the average intensity outside the periphery). This value can then be used to classify the detected circle as an imaged bubble or liquid droplet using one or more classification techniques (e.g., classification thresholds, naive Bayesian inference, machine learning-based classifiers, etc.).

[0267] In some embodiments, detection of an air bubble in outlet channel 2239 can indicate a fault condition in the device or that the unpolymerized mixture is nearly depleted. Thus, in some embodiments, in response to detection of an air bubble in outlet channel 2239, the pressure of the fluid stream corresponding to the channel transporting the unpolymerized mixture (e.g., channel 2141 shown in Figures 26B and 26C) can be reduced to prevent continued bubble formation and / or air flow into the outlet channel. In situations where the pressure is not reduced sufficiently, an undesirable situation (sometimes referred to as a "blowout scenario") can occur in which air flow enters the outlet channel. Image 2200D shown in Figure 27D illustrates an example of a blowout scenario in which air flow 2212 is being forced into outlet channel 2239.

[0268] FIG. 27E shows image 2200E in which no droplets or bubbles appear in outlet channel 2239. As a result, no circles were detected in image 2200E. In some embodiments, such a scenario may indicate a fault condition. For example, image 2200E may occur when the pressure of the unpolymerized mixture fluid flow is too low or too high to support stable droplet generation. In these scenarios, outlet channel 2239 may be completely filled with either an immiscible fluid (e.g., if the pressure of the unpolymerized mixture fluid flow is too low) or an unpolymerized matrix (if the pressure of the unpolymerized mixture fluid flow is too high). In some embodiments, in response to not detecting any circles in outlet channel 2239, the device's controller (e.g., controller 2124 of device 2100) may implement a fault recovery routine to resume stable droplet generation. Actions taken by the controller, including fault recovery routines, are described in further detail herein.

[0269] 28 shows an exemplary process 2300 for controlling the pressure of a fluid flow in a droplet generating device, for example, to control the size of droplets generated. At least some operations of process 2300 may be performed by a microfluidic system, such as device 2100 or a portion thereof (e.g., controller 2124). In some embodiments, one or more operations of process 2300 may be performed by one or more remote computing systems external to device 2100.

[0270] Operations of process 2300 include acquiring 2302 one or more images. For example, the images may be captured by a camera, such as optical sensor 2132 of device 2100. The images may correspond to images 2200A-2200E described in connection with FIGS. 27A-27E. Similar to images 2200A-2200E, the acquired images may be framed to include an extended region of an outlet channel (e.g., outlet channel 2139), which may potentially include droplets and / or bubbles. In some implementations, after acquiring 2302 one or more images, gamma correction may be applied to one or more images. Gamma correction may convert pixel array data acquired from the camera into a numerical value proportional to actual photon intensity, thereby optimizing bit usage, for example, by accounting for the nonlinear way humans perceive light and color.

[0271] The operations of process 2300 may also include detecting edges in the images (2304) (possibly after applying gamma correction to one or more images) and detecting circles based on the detected edges (2306). Both of these operations may be performed by an image processing module implemented in a controller of the microfluidic device (e.g., controller 2124 of device 2100). In some embodiments, operations 2304, 2306 may be performed on a remote computing device external to device 2100. Detecting edges (2304) may include implementing an edge detection algorithm, such as a Canny edge detector, a Gaussian edge detector, or a threshold-based edge detector. In some embodiments, detecting edges (2304) may also include using asymmetric filtering to mitigate bias caused by the contours of channels in the microfluidic device. For example, in one or more images, the contours of a chamber may appear, at least in part, as two dark vertical lines in the image (as appearing in Figures 27A-27E). In such cases, the filter can be intentionally biased to be more sensitive to vertical pixel intensity variations by using a difference-of-Gaussians kernel along the vertical axis, or by using the kernel [[0.5,1,0.5], [0,-4,0], [0.5,1,0.5]]. Various filters can be selected and used depending on the specific optical and droplet characteristics of the system. Detecting circles based on detected edges (2306) can be performed using one or more shape detection algorithms, such as a Hough transform. In some implementations, one or more images can be reduced in size (e.g., by downsampling) before detecting edges in the image (2304) and / or before detecting circles based on detected edges (2306). For example, one or more images can be downsampled (e.g., digitally downsampled) to between one-quarter and one-half of their original resolution (e.g., 25% of the original resolution, 30% of the original resolution, 40% of the original resolution, 50% of the original resolution, etc.).This downsampling may have the advantage of increasing the speed of edge detection and / or circle detection. For example, the number of operations involved in performing a Hough transform (an exemplary algorithm that may be used to detect circles) may increase proportionally to the fourth power of the image size. Thus, downsampling an image before using a Hough transform to perform circle detection may result in significant time savings (e.g., allowing for 5-15 times faster performance of the Hough transform and 2-10 times faster performance of the overall process (2300)). For example, in an embodiment where an original image having a resolution of 100x100 pixels was downsampled to an image having a resolution of 50x50 pixels before performing the Hough transform, the Hough transform performed approximately 10 times faster and process 2300 performed approximately 6 times faster compared to an embodiment without downsampling. In embodiments where downsampling is performed, after detecting edges in the image (2304) and / or detecting circles based on the detected edges (2306), the resulting image data may be upscaled. For example, the image may be upscaled by a factor ranging from 2 to 4. In some cases, after selecting a magnification factor to detect edges in the image (2304) and / or detect circles based on the detected edges (2306), the downsampled image can be restored to its original size.

[0272] If no circle is detected in operation 2306, process 2300 may include generating (2308) a signal to adjust the pressure of at least one fluid stream according to a fault recovery routine. For example, the one or more images may be similar to image 2200E shown in FIG. 27E without the detected circle. In some embodiments, a controller of the microfluidic device (e.g., controller 2124 of device 2100) may control one or more pumps (e.g., pump 2126) to control the flow rate (e.g., by controlling the pressure) of a fluid stream within device 2100. In some embodiments, this fluid stream may correspond to the flow of the unpolymerized mixture through the device. However, in other embodiments, both the flow of the unpolymerized mixture and the pressure of the immiscible fluid stream may be adjusted.

[0273] The fault recovery routine may include implementing a control scheme for controlling the flow rate of the fluid stream (e.g., by controlling pressure), where the control scheme operates without using feedback control. For example, the fault recovery routine may include implementing a simple control scheme that gradually increases the pressure of the fluid stream of the unpolymerized mixture until stable droplet generation begins or until a maximum pressure value is reached. In some embodiments, the fault recovery routine may include implementing a simple control scheme that gradually decreases the pressure of the fluid stream of the unpolymerized mixture until stable droplet generation begins or until a minimum pressure value is reached. In some embodiments, the fault recovery routine may include implementing a simple control scheme that continuously varies the pressure of the fluid stream of the unpolymerized mixture between a minimum pressure value and a maximum pressure value until stable droplet generation begins. Using a simple static controller instead of feedback control in the presence of a fault condition can prevent undesirable or unstable behavior of the feedback controller when droplets are not detected or when droplets of extreme sizes are detected.

[0274] Alternatively, if circles are detected in act 2306, process 2300 may include determining (2310) for each circle whether one or more filtering conditions are met. For example, a captured image with a detected circle may correspond to image 2200A, 2200B, or 2200C. The one or more filtering conditions may include criteria for excluding circles to prevent false positives or low-quality circle data from biasing the size estimates.

[0275] In some implementations, the filtering conditions may include an indication that the detected circle is based on the detected edges of multiple droplets. For example, this may result from erroneously identifying a space between multiple droplets as a detected circle. In some implementations, the indication that the detected circle is based on the detected edges of multiple droplets may result from a determination that the detected circle is substantially smaller than one or more other detected circles in the same image. This indication may also result from a determination that the detected circle shares an edge with one or more other detected circles of substantially larger size.

[0276] In some implementations, the filtering conditions may include an indication that a detected circle overlaps at least one additional circle in the same image. While droplets may occasionally touch each other in an outlet channel (e.g., outlet channel 2139), they typically do not overlap. Thus, the detected circle should also not overlap in the imaged outlet channel (e.g., outlet channel 2239). Thus, receiving an indication that a detected circle overlaps at least one additional circle may suggest that at least some of the overlapping circles may be false positives.

[0277] In some embodiments, the filtering criteria may include unexpected detected signals within the perimeter of the detected circle. In some embodiments, detected signals (e.g., dark pixels) in one or more images are expected to be present only at edges corresponding to droplets, bubbles, and / or the microfluidic device itself. In some embodiments, certain types of detected signals (e.g., faint signals or signals of particular sizes and / or shapes) corresponding to tumor-like masses or other cells located within the detected droplets may also be expected. However, if unexpected signals that differ from these expected signals are detected within the perimeter of the detected circle, this may suggest that the detected circle may be a false positive or is otherwise abnormal. In some embodiments, the unexpected detected signals may correspond to detected signals arising from within the perimeter of the detected circle, where the detected signals exceed a threshold signal level.

[0278] In some implementations, the filtering criteria may include determining that the detected circle is within a proximity threshold from the wall of the imaged device (sometimes referred to herein simply as the "wall"). Droplets located on the device wall may compress against the wall and lose their spherical shape. This may reduce the performance of the circle detection algorithm and / or bias the estimate of the droplet size. Furthermore, because the device wall may appear in the image as high-contrast edges, there is a risk that these edges may be erroneously identified as part of the detected circle, which may lead to false positives. Therefore, in some implementations, it may be beneficial to filter out detected circles that are too close to the device wall.

[0279] If any of the filtering conditions are met at operation 2310 for a particular detected circle, process 2300 may include filtering out the detected circle 2312. As discussed above, operation 2312 may result in a more accurate estimate of droplet size by filtering out false positives and low-quality detected circles.

[0280] If none of the filtering conditions are met in operation 2310, process 2300 may include identifying pixels corresponding to the detected edge located within a threshold distance from the perimeter of the detected circle (2314) and calculating an average distance from the center of the detected circle to at least a portion of the identified pixels (2316). This calculated average (which may be a weighted average) may be used as the updated radius of the detected circle. In some implementations, a metric other than the average may be calculated, provided that the metric is derived from and representative of the distance from the center of the detected circle to at least a portion of the identified pixels. Compared to single-pass circle detection algorithms that detect circles with a radial resolution of one pixel or greater, this multi-pass approach to estimating the size of the detected circle provides sub-pixel radial resolution. This resolution is particularly important for microfluidic applications, where droplets may have radii in the range of only 5-20 pixels in some cases (although in other examples, the radii may be up to 50 pixels, 100 pixels, 250 pixels, etc.). A visual representation of this process is provided in Figure 29 and is described in further detail below.

[0281] After detecting a high-quality circle and estimating its size with sub-pixel radial resolution, process 2300 may include determining 2318 whether the detected circle corresponds to a liquid droplet or an air bubble. As previously mentioned, with respect to Figures 27A-27C, an air bubble may be distinguished from a liquid droplet based on the darkness of pixels corresponding to the edge of the detected circle, with an air bubble (e.g., detected circle 2210) having darker edges than liquid droplets (e.g., detected circles 2202, 2204, 2206, 2208).

[0282] If the detected circle is an air bubble, process 2300 may include reducing (2320) the pressure of at least one fluid stream. For example, operation 2320 may include reducing the pressure (and therefore the flow rate) of a fluid stream corresponding to a channel for transporting the unpolymerized mixture (e.g., channel 1241 shown in Figures 26B and 26C). In some embodiments, the fluid flow may be stopped entirely and / or the droplet generation device may be powered off. As mentioned above, this may prevent continued bubble formation and / or air flow into the device's outlet channel (e.g., outlet channel 2139). Operation 2320 may also prevent loss or waste of unpolymerized mixture (e.g., sample) under operating conditions where droplets are not properly generated. Operations 2300 and 2320 may further provide the advantage of automatically detecting when sample is depleted (resulting in bubble formation), thereby eliminating the need for the droplet generation device to store and track the amount of available sample remaining.

[0283] If the detected circle is a droplet, process 2300 may include determining whether the standard deviation (or any other measure of dispersion) of the estimated droplet size(s) in the image is below a threshold (2322). In some implementations, a large standard deviation of the estimated droplet size(s) in the image may indicate a fault condition, since continuously generated droplets should not typically vary substantially in size. As a result, if the standard deviation exceeds the threshold, process 2300 may include generating a signal (2324) to make an adjustment to the pressure (and flow rate) of at least one fluid flow in accordance with a fault recovery routine. The fault recovery routine may be substantially similar to the exemplary fault recovery routine described above.

[0284] In some implementations, if the estimated sizes of droplets identified in a single image vary by more than a threshold amount, the largest estimated size may be treated as most representative of the actual droplet size. For example, smaller estimated sizes may be discarded, or a weighted average of all estimated sizes may be taken, with smaller estimated sizes being given less weight than larger estimated sizes. This may prevent false positive detections of small droplets, which may occur more frequently than false positive detections of large droplets due to noise and / or other factors.

[0285] If the standard deviation (or any other dispersion) of the estimated droplet size(s) in the image is below a threshold, process 2300 may include comparing (2326) the estimated droplet size(s) to a target size. For example, the target size for the droplets may be between 220 microns and 300 microns. In some implementations, the target size may be manually adjusted by a user of the device. The comparison of the estimated droplet size(s) to the target size may be used to calculate an error signal.

[0286] Operations of process 2300 may also include generating 2328 a signal to make an adjustment to the pressure (and flow rate) of at least one fluid stream using feedback control. For example, as described above, the pressure may be adjusted using a device controller (e.g., controller 2124 of device 2100) to control one or more pumps (e.g., pump 2126) to control the pressure of the fluid stream within device 2100. This fluid flow may correspond to the flow of unpolymerized mixture through the device (e.g., channel 2141). Feedback control may be implemented using an error signal calculated based on a comparison of the estimated droplet size(s) in operation 2326 to a target size. In some cases, the error signal or estimated droplet size(s) may be averaged over multiple past images (e.g., three images, five images, ten images, twenty images, etc.) to reduce noise. Feedback control may include proportional control, integral control, derivative control, or any combination of the above. We use the term "proportional control" broadly to include control system techniques based on a response proportional to an error signal. The error signal can be the difference between a desired process value (or setpoint) and the current value of a controlled process variable (e.g., the size of a detected droplet). We use the term "integral control" broadly to include control system techniques based on a response proportional to the integral (e.g., the time integral) of the error signal. We use the term "derivative control" broadly to include control system techniques based on a response proportional to the derivative (e.g., the derivative with respect to time) of the error signal.

[0287] Once feedback control is initiated, it may continue until either an air bubble is detected or a fault condition is identified (e.g., no circle is detected in the captured image or the standard deviation of the estimated droplet size(s) in the image exceeds a threshold). To enable this continuous control, process 2300 may be repeated for multiple images captured by a camera (e.g., optical sensor 2132) of the microfluidic device. For example, in some embodiments, the multiple images may be frames of a single captured video.

[0288] Figure 29 shows in more detail a process for estimating droplet size, which corresponds to operations 2302, 2304, 2306, 2310, 2312, 2314, 2316 of process 2300 shown in Figure 28. All of these operations may be performed by an image processing module (e.g., an image processing module of a microfluidic device) and may be implemented in a controller (e.g., controller 2124).

[0289] Imaged object 2402 represents a magnified view of the original image (e.g., captured in operation 2302 of process 2300) and focuses on a single droplet. In operation 2404, edge detection (corresponding to operation 2304 of process 2300) may be performed to generate detected edges 2406 corresponding to imaged object 2402. Imaged object 2402 may also be enhanced (in operation 2416) to result in enhanced object 2418. For example, imaged object 2402 may be enhanced using conventional feature enhancement or edge sharpening techniques, such as an edge sharpening filter or a "difference of Gaussians" process, to increase the visibility of edges in the imaged object. In one example, a "difference of Gaussians" process may involve calculating the difference between two Gaussian-blurred versions of an imaged object (e.g., a first version blurred using a Gaussian function with a standard deviation of 2 pixels, and a second version blurred using a Gaussian function with a standard deviation of 1 pixel, although other parameter values ​​may be available and preferable depending on the hardware and lighting conditions used to acquire the images). In this particular example, the resulting image may preserve visual (e.g., spatial) information (such as edges of the imaged object) with frequencies in the range of approximately 1 pixel to 3 pixels, while blurring or removing other features.

[0290] The detected edge 2406 may be used as input to a circle detection process (at act 2408) to obtain a detected circle 2410 having a radius r. Act 2408 corresponds directly to act 2306 of process 2300 and may similarly be implemented using a Hough transform. In some implementations, the detected circle 2410 may have a radial resolution of one pixel or greater.

[0291] As discussed above in connection with FIG. 28 , in some implementations, the original image including the imaged object 2402 may be reduced in size (e.g., by downsampling) before edge detection (operation 2404) or circle detection (operation 2408). For example, one or more images may be downsampled (e.g., digitally downsampled) to one-quarter to one-half of their original resolution (e.g., 25% of the original resolution, 30% of the original resolution, 40% of the original resolution, 50% of the original resolution, etc.). This downsampling may have the advantage of increasing the speed of edge detection and / or circle detection. In implementations in which downsampling is performed, the resulting image data may be enlarged after edge detection (operation 2404) and / or circle detection (operation 2408). For example, the image may be enlarged by a factor ranging from 2 to 4. In some cases, the enlargement factor may be selected to restore the downsampled image to its original size.

[0292] At decision point 2412 (corresponding to operation 2310 of process 2300), the image processing module may determine whether one or more filter conditions are satisfied. As described above in connection with operation 2312 of FIG. 28 , if the filter conditions are satisfied, the detected circle 2410 may be excluded 2414 from the droplet size estimation. However, if the filter conditions are not satisfied, the detected circle 2410 may be processed along with the enhanced object 2418 to calculate 2420 a representative radius based on all pixels in the enhanced object 2418 having a radius of r+d to rd, where d represents a threshold distance from the perimeter of the detected circle 2410. In some implementations, the calculation of the representative radius may be based on only a portion of these pixels. Operation 2420 corresponds to operations 2314, 2316 of process 2300 and has the similar advantage of providing sub-pixel radial resolution as described above. The representative radius calculated in operation 2420 may be an average of the distances, a weighted average of the distances, or another metric that is derived from and representative of the distances from the centers of the detected circles to at least a portion of the identified pixels. For example, the representative radius may be calculated using the following formula:

[0293]

number

[0294] During the ceremony, w i is the weight value reflecting the intensity value of pixel i, and r i where π reflects the distance of pixel i from the center of the detected circle, and n represents the total number of pixels included in the calculation. While various alternative reference values ​​can be used, the calculation of the representative radius described by the above equation may have the advantage of placing a higher weight on radii corresponding to pixels with greater intensity. The calculated representative radius from operation 2420 may then be used to update the radius of the detected circle 2410 for further processing (e.g., as described in connection with process 2300).

[0295] 30 and 31, experimental data demonstrating the capabilities and advantages of the techniques described herein are presented.

[0296] FIG. 30 shows graph 2500 from an experiment in which a user manually set a target size or “setpoint size” for droplets generated by a microfluidic device, increasing the target size at approximately 180 seconds and decreasing the target size at approximately 260 seconds. Note, however, that in some embodiments, the setpoint may be automatically set (and changed) by a computer system based, for example, on one or more characteristics of the sample, one or more characteristics of the droplet generating device, etc. The setpoint size (corresponding to the target radius of the droplets) is shown by trace 2502 and is expressed in pixels. Trace 2504 corresponds to the estimated droplet size for each captured image (e.g., each frame of a video captured by a camera in the microfluidic device), also expressed in pixels. Trace 2506 is a running average of trace 2504 to reduce noise effects, also expressed in pixels. Trace 2508 shows the pressure applied to the holding chamber of the unpolymerized mixture to control the flow rate of the unpolymerized mixture. The pressure is directly controlled by the microfluidic device's controller and is measured in millibars. Trace 2510 shows the pressure applied to the holding chamber for the immiscible fluid (held constant in this experiment) and is measured in millibars. Trace 2512 shows the circle counts detected in each captured image.

[0297] For the first 50 seconds, stable droplet generation had not yet begun, as evidenced by many image frames in which count (trace 2512) and size (trace 2504) were at zero. In response to the lack of droplet detection during the first 50 seconds, as described above, the applied pressure on the unpolymerized mixture stream was gradually increased to a maximum pressure of 800 mbar and held constant until stable droplet generation began. Once stable droplet generation began (approximately 50 seconds), feedback control of the unpolymerized mixture stream pressure (trace 2508) successfully enabled the estimated size traces (trace 2504 and trace 2506) to track the setpoint size (trace 2502). Thus, graph 2500 demonstrates the successful use of pressure control to control the fluid flow rate within a microfluidic device and thereby the size of droplets generated.

[0298] FIG. 31 shows graph 2600 from a second experiment, in which the target size or “setpoint size” of droplets generated by the microfluidic device was manually set by a user and varied over time. The traces in graph 2600 are similar to those in graph 2500. The setpoint size (corresponding to the target radius of the droplets) is shown by trace 2602 and is measured in pixels. Trace 2604 corresponds to the estimated droplet size for each captured image (e.g., each frame of a video captured by the microfluidic device camera), also measured in pixels. Trace 2606 is a running average of trace 2604 to reduce noise effects, also measured in pixels. Trace 2608 shows the pressure applied to the holding chamber of the unpolymerized mixture to control the flow rate of the unpolymerized mixture. The pressure is directly controlled by the microfluidic device controller and is measured in millibars. Trace 2610 shows the pressure applied to the holding chamber for the immiscible fluid (held constant in this experiment), measured in millibars, and trace 2612 shows the circle counts detected in each captured image.

[0299] Similar to graph 2500, the overall trend of estimated size (trace 2606) tracks fluctuations in setpoint size (trace 2602). However, in some instances, fault conditions 2620 occurred where no droplets were detected, and the count trace (trace 2612) and estimated size trace (trace 2604) correspondingly dropped to zero. While this behavior is not ideal, the experimental data shown in graph 2600 demonstrates the ability of the techniques described herein to recover from these fault conditions 2620. As discussed above in connection with FIG. 28, as each fault condition 2620 was identified, the microfluidic device's controller replaced the use of pressure feedback control with a simple pressure controller fault recovery routine that gradually increased, decreased, or maintained the pressure applied to generate the flow of unpolymerized mixture (trace 2608) until stable droplet generation resumed. In each instance of fault condition 2620, droplet generation could be successfully resumed without further user intervention.

[0300] While the experiments shown in connection with Figures 30 and 31 demonstrate dynamic control of pressure applied to only a single fluid stream (i.e., the unpolymerized mixture stream), the present disclosure is not intended to be limiting. As previously discussed, droplet size can depend on the pressure (and thus flow rate) of both the unpolymerized mixture stream and the immiscible fluid stream. As a result, in some embodiments, in addition to, or instead of, controlling the pressure (and flow rate) of the unpolymerized mixture stream, the pressure (and flow rate) of the immiscible fluid stream can also be controlled. For example, in some embodiments, simultaneously controlling the pressure of both fluid streams in a microfluidic device can enable additional control of the droplet generation process, including control of the rate of generation and the size of the droplets generated.

[0301] 32 shows an exemplary process 2700 for controlling the size of droplets generated by a device (e.g., a microfluidic device such as device 2100). The operations of process 2700 may be performed by a microfluidic system such as device 2100 or a portion thereof (e.g., controller 2124). In some embodiments, one or more operations of process 2700 may be performed by one or more remote computing systems external to device 2100.

[0302] Operations of process 2700 may include acquiring 2702 one or more images representing a flow path in a microfluidic system that facilitates interaction between a first fluid stream and a second fluid stream. Acquiring the one or more images may include capturing images with an optical sensor (e.g., optical sensor 2132), which may in some cases be a camera. The one or more images include an image of at least one droplet generated in a flow path in a region of the device, wherein the at least one droplet is not compressed by one or more walls of the device. For example, the at least one droplet may be imaged within an expanded region of an outlet channel (e.g., outlet channel 2139) of the device.

[0303] The operations of process 2700 also include processing one or more images to identify at least one droplet generated in the flow path by interaction between the first and second fluid streams (2704). The first fluid stream may include a flow of a hydrophilic solution (e.g., an unpolymerized mixture containing a fluid matrix material and cells), and the second fluid stream may include a flow of a hydrophobic solution (e.g., an immiscible fluid such as oil). Processing the one or more images may include detecting an edge of at least one droplet in at least one of the one or more images (e.g., using a Canny edge detector) and identifying a first set of pixels corresponding to the detected edge of the at least one droplet. For example, the first set of pixels may be a circle representation generated based on the detected edge of the at least one droplet (e.g., using a Hough transform). Processing the one or more images to identify the at least one droplet may also include down-scaling the one or more images before detecting the edge of the at least one droplet and / or before identifying the first set of pixels corresponding to the detected edge of the at least one droplet. For example, as described above, one or more images may be downsampled (e.g., digitally downsampled) to one-quarter to one-half of their original resolution (e.g., 25% of the original resolution, 30% of the original resolution, 40% of the original resolution, 50% of the original resolution, etc.). In embodiments where downsampling is performed, processing the one or more images to identify at least one droplet may further include magnifying the one or more downsampled images after detecting the edge of the at least one droplet and / or after identifying a first set of pixels corresponding to the detected edge of the at least one droplet. For example, the images may be magnified by a factor ranging from 2 to 4. In some cases, the magnification factor can be selected to restore the downsampled image to its original size. Processing the one or more images may also include identifying a second set of pixels located within a threshold distance from the first set of pixels and calculating an average distance of at least a portion of the second set of pixels from a predetermined location within the at least one droplet.In some examples, the predetermined location within the at least one droplet may be the center of the at least one droplet (or the center of a circular representation of the at least one droplet). In some implementations, identifying the second set of pixels may occur after any shrinking and re-scaling of the one or more images being processed. Processing the one or more images may further include excluding data corresponding to a detected edge of the at least one droplet if one or more filtering conditions are met. The filtering conditions may include an indication that the first set of pixels corresponds to a detected edge of multiple droplets, an indication that the first set of pixels overlaps with at least one additional set of pixels (e.g., pixels corresponding to another circular representation of another droplet), a detected signal that meets a threshold signal level condition and originates from within the perimeter of the first set of pixels, and / or a determination that the first set of pixels is less than a proximity threshold from a wall of the imaged device.

[0304] Operations of process 2700 also include estimating 2706 a size of at least one droplet and determining 2708 whether the size of the at least one droplet satisfies a threshold condition. Estimating the size of the at least one droplet may include estimating the size based on an average distance of at least a portion of a set of pixels (e.g., the second set of pixels described above) from a predetermined location within the at least one droplet. In some implementations, estimating the size of the at least one droplet may include estimating the size with a sub-pixel radial resolution. Determining whether the size of the at least one droplet satisfies a threshold condition may include comparing the size of the at least one droplet to a target size obtained by user input (e.g., to generate an error signal).

[0305] Operations of process 2700 also include generating 2710 a signal to make an adjustment to the pressure of at least one of the first fluid stream or the second fluid stream in response to determining that the size of the at least one droplet satisfies a threshold condition. The signal may be configured to increase or decrease the pressure of at least one of the first fluid stream or the second fluid stream based on the size of the at least one droplet. In some implementations, generating the signal may include generating the signal using proportional control, integral control, and / or differential control using a feedback controller. In some implementations, generating the signal may include generating the signal without feedback control (e.g., using a simple static controller) if no droplets are identified in one or more images and / or if the standard deviation of the size of at least one droplet (e.g., two or more droplets) exceeds a threshold. In some implementations, generating the signal may be based on estimated sizes of multiple droplets in the same or different captured images (e.g., consecutive frames of a video).

[0306] Additional operations of process 2700 may include: In some embodiments, process 2700 may include processing one or more images to identify air bubbles in the flow path; In some embodiments, process 2700 may include storing the one or more images and / or data representing the estimated size of the at least one droplet in a storage device; In some embodiments, process 2700 may include transmitting the one or more images and / or data representing the estimated size of the at least one droplet to a remote computing device.

[0307] Liquid level detection using light reflection

[0308] Accurate and precise measurement and monitoring of liquid levels in containers (e.g., tubes) is important in many applications, including chemical or biological analysis and medical diagnostics, for example, in the context of the MOS generation systems and methods described above. For example, the liquid level sensing approaches described herein can be used in combination with one or more aspects of the MOS generation systems and methods described above, including in combination with approaches to closed-loop control and / or droplet size determination described herein. In some examples, liquid levels are sensed in these MOS generation systems and methods using optical reflection, such as total internal reflection (TIR). Optical reflection approaches for liquid level sensing have advantages over other liquid level sensing approaches, such as capacitive, ultrasonic, or pressure-based approaches, in that optical-based approaches are effective, inexpensive, and accurate.

[0309] The approach described herein for liquid level sensing uses total internal reflection (TIR). Liquid level sensing is based on incorporating an optical interface within a container (e.g., a tube), where the optical interface is between the medium (air or liquid) within the container and the inner surface onto which light is incident. When the liquid level within the container is below the optical interface, TIR occurs, and light incident on the optical interface is completely reflected back by the optical interface; when the liquid level is above the optical interface, there is no TIR, and light incident on the optical interface is substantially (or completely) transmitted through the liquid and / or container. The liquid level can be determined based on the results of measuring the reflected and / or transmitted light.

[0310] In some embodiments, the liquid level sensor includes a container having at least one inner surface (e.g., for an optical interface) and a pair of a light source (e.g., a light emitting diode (LED) or laser diode) and a photodetector (e.g., a photodetector, photodiode, or phototransistor). The pair of light source and photodetector may be closely spaced apart and packaged as an electro-optical package. The light source and photodetector may be located on the same side of the container and positioned adjacent to the container for measurement (e.g., approximately a certain distance apart, e.g., 0 mm, 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, or any suitable distance). Light from the light source passes through the outer surface of the container, enters the at least one inner surface, is reflected by the at least one inner surface, and is detected by the photodetector.

[0311] In some embodiments, the light source and light detector pair can be moved along the length of the container to measure the liquid level within the container. In some embodiments, multiple light source and light detector pairs can be positioned at a series of positions along the length of the container, and the liquid level within the container can be determined based on multiple measurements of the multiple light source and light detector pairs.

[0312] The container in which the liquid level is sensed is a structure that defines a space for holding the fluid medium to be measured. The container may be a reservoir, vessel, tube, canister, tank, bottle, or any suitable structure.

[0313] In some embodiments, the container includes a protrusion extending along the length of the container. In some embodiments, the protrusion and container can be formed separately, and then the protrusion is attached to the container (e.g., via an index-matching material). In some embodiments, the container and protrusion can be integrally formed, for example, by molding or 3D printing, such as stereolithography (SLA) printing. The container and / or protrusion can be optically transparent or at least partially transparent. The container material can be glass, plastic, or polymer, such as polypropylene, polyethylene, polystyrene, polycarbonate, polymethyl methacrylate (PMMA), acrylonitrile butadiene styrene (ABS), high-clarity photopolymer Somos Watershed or Waterclear resin, or any substantially transparent resin with a highly polished surface that can transmit light with minimal diffusion. The container can be custom finished with a highly polished surface.

[0314] The protrusion can include at least one interior surface. For example, the protrusion can be a 90° V-shaped raised rib on the body of the container. The body defines a space for holding a liquid. The protrusion can include first and second sides that define a 90-degree angle, and a third side that is part of the body of the container. The protrusion can be configured for retroreflection of light.

[0315] In some embodiments, light may enter the container along the horizontal direction, strike a first inner surface of a first side at an angle of incidence (e.g., 45°), be reflected by the first inner surface toward a second inner surface of a second side of the protrusion (e.g., at a 45° angle of incidence), and then be reflected by the second inner surface back along the horizontal direction toward the photodetector. When the liquid level is lower than the incident position of the light on the first and second inner surfaces of the protrusion, the incident angle is greater than the critical angle (defined by the refractive index of the material of the protrusion and the refractive index of air) of each of the first and second inner surfaces, and total internal reflection occurs at the interface between the air and the protrusion. When the liquid level is higher than or equal to the incident position of the light on the first inner surface, for example, when the light enters the liquid, the incident angle is smaller than or equal to the critical angle (defined by the refractive index of the material of the protrusion and the refractive index of the liquid), and total internal reflection does not occur because the refractive index of the liquid (e.g., 1.3) is closer to the refractive index of the material of the protrusion (e.g., 1.5) than to that of air (e.g., 1.0). The light then exits into the liquid and transmits through the container, and the photodetector receives much less or no reflected light. In some cases, if the refractive index of the liquid is close to the refractive index of the material of the protrusion (e.g., plastic), the critical angle can be large and total internal reflection occurs only when light hits the protrusion at a very small grazing angle.

[0316] These liquid level sensing technologies can address existing challenges for liquid level sensing. For example, the liquid level sensing technologies can improve the accuracy of liquid level measurements, thereby reducing variability in biological or chemical processes, leading to higher product quality, reduced costs, and reduced waste. These technologies can provide cost-effective, compact, and easily fabricated liquid level sensors or systems. These technologies can provide precise and reliable liquid level measurement sensors and systems that can meet the demands of highly complex automated processing systems, tighter process control needs, and increasingly strict regulatory environments. The technologies can also provide non-contact liquid level sensing through the walls of a sealed container, so that neither the liquid level sensor nor the liquid is contaminated. The technologies described herein can be applied to any fluid-related process that requires a precise amount of liquid in a container. The technologies described herein can be used in many applications, such as chemical / biological analysis and medical diagnostics. In addition to liquids, the technologies can be applied to any other type of fluid medium, such as any medium with a refractive index higher than that of air. The liquid can include one or more different types of fluid medium.

[0317] 33A-33C show an example of a vessel 3100 for liquid level detection. FIG. 33A is a schematic diagram of the vessel 3100, FIG. 33B is a side view A-A' of the vessel 3100, and FIG. 33C is a cross-sectional view B-B' of the vessel 3100. The vessel 3100 includes a protrusion with at least one inner surface as an optical interface for total internal reflection, which can be used for liquid level detection. For example, the vessel may be a sample or effluent vessel used in the MOS generation system described above. For example, if liquid levels are detected in both the sample and effluent vessels, the initial and / or final volume of liquid in the input vessel may be determined, and the volume of sample or waste in the effluent vessel may be determined. In some cases, the difference between the volume of liquid in the input vessel and the volume of sample or waste in the effluent vessel may be obtained for further analysis.

[0318] As shown in Figures 33A-33C, the container 3100 includes a body 3110 extending along a longitudinal direction 3101 from a bottom 3102 to a top 3104. The body 3110 defines a space for holding a liquid, such as water, a solution, an oil, or any suitable fluid medium. The bottom 3102 of the container 3100 may have a flat surface, a conical shape, or any suitable shape. The top 3104 may include a cover having a recess or groove for sealing.

[0319] The container 3100 may be configured to have a liquid holding volume of, for example, 1 milliliter (mL), 2 mL, 5 mL, 10 mL, 20 mL, 50 mL, 100 mL, 200 mL, 500 mL, 1 liter (L), 2 L, 5 L, 10 L, 100 L, or any suitable volume. Different liquid levels within the container 3100 may correspond to different volumes of liquid held within the container 3100. In some embodiments, the container 3100 includes a series of volume level indicia corresponding to a series of positions on the body 3110 along the longitudinal direction, each volume level indicia corresponding to a respective volume.

[0320] In some embodiments, as shown in Figures 33A-33C, the container 3100 has a cylindrical shape with a circular cross-section. The container may also have any other suitable shape, such as a rectangular parallelepiped shape with a rectangular cross-section. In some embodiments, the container 3100 may have equal sizes (e.g., diameters) along its length, or may have variable sizes that vary from small to large or large to small along its length. For illustrative purposes only, a container having a cylindrical shape with a circular cross-section is used as an example for the description herein.

[0321] In some embodiments, as shown in Figures 33A-33C, the container 3100 includes a protrusion 3120 that can protrude inward from the body 3110 of the container 3100. The protrusion 3120 can extend along the longitudinal direction 3101 of the container 3100, for example, from the bottom 3102 to the top 3104, as shown in Figure 33B. The protrusion 3120 can include a first side surface 3122, a second side surface 3124, and a third side surface 3126. Figure 33C shows a cross-section B-B', in which the cross-section of the protrusion 3120 includes a first inner surface 3132 on the first side surface 3122, a second inner surface 3134 on the second side surface 3124, and a third surface 3136 on the third side surface 3126. As will be discussed in more detail below, the protrusion 3120 is configured so that light entering from the outer surface 3112 of the main body 3110 (e.g., the surface open to the external environment) can be totally reflected internally by the first inner surface 3132 of the first side toward the second inner surface 3134 of the second side, and then totally reflected internally by the second inner surface 3134 back out of the container 3100.

[0322] In some embodiments, the protrusion 3120 is a V-shaped rib, as shown in FIGS. 33B and 33C. The first side 3122 and the second side 3124 may be connected together at an edge 3123, and the first inner surface 3132 and the second inner surface 3134 may be connected at a corner 3133. The edge 3123 may be a straight line along the longitudinal direction 3101, and the corner 3133 may be a point. In some embodiments, the edge 3123 may be any other suitable shape, for example, an arc or curved shape, and thus the corner 3133 may have an arc or curved shape. In some embodiments, the first inner surface 3132 and / or the second inner surface 3134 are straight lines, as shown in FIG. 33C. The angle defined by the first inner surface 3132 and the second inner surface may be substantially equal to 90 degrees. In some embodiments, the first inner surface 3132 and / or the second inner surface 3134 can have at least a partially arcuate or curved shape.

[0323] The body 3110 and the protrusions 3120 may be optically transparent or at least partially transparent so that light can propagate through the body 3110 and the protrusions 3120. In some embodiments, the body 3110 of the container 3100 may be made of a first material, such as glass, plastic, or a polymer. The protrusions 3120 may be made of a second material, such as glass, plastic, or a polymer. In some embodiments, the second material is the same as the first material. The third side 3126 of the protrusions 3120 may be part of the body 3110 and may be integral with the remainder of the body 3110. In some embodiments, the second material is different from the first material. The third side 3126 of the protrusions 3120 may be attached or glued to the inner surface of the body 3110.

[0324] 34A-34B show an example of a liquid level sensor 3200 for measuring the liquid level in the container 3100 of FIGS. 33A-33C when the liquid level is below (FIG. 34A) or above (FIG. 34B) the light incidence position, respectively, in accordance with one or more embodiments of the present disclosure. The liquid level sensor 3200 includes a container 3100 having a protrusion 3120 and a sensing pair 3210 of a light source 3212 and a light detector 3214.

[0325] The protrusion 3120 may be a V-shaped rib having an angle Φ defined by a first inner surface 3132 and a second inner surface 3134. Light emitted from the light source 3212 enters through the outer surface 3112 of the body 3110 and is incident at an incident angle θ i1 may be incident on the first inner surface 3132 at

[0326] According to Snell's law, the critical angle for total internal reflection (TIR) ​​is sin -1 (n i / n o ) where n i is the refractive index of the protrusion 3120, and n ois the refractive index of the medium in contact with the first inner surface 3132 of the container 3100. As shown in FIG. 34A, when the liquid level is lower than the entrance position, the medium in contact with the first inner surface 3132 has a refractive index n that is the same as that of the gas 3202, e.g., 1. o As shown in FIG. 34B, when the liquid level is higher than or equal to the incident position, the medium in contact with the first inner surface 3132 has a refractive index n that may be the same as that of the liquid 3204, e.g., 1.3. o The material of the protrusion 3120 is an aqueous solution having a refractive index n that can be greater than 1.3 and less than 2.0. i It has.

[0327] For purposes of example only, the refractive index n of the material of the protrusion 3120 i is set equal to 1.5. Thus, when the liquid level is lower than the incident position, n0=1 and the TIR critical angle is 41.8°, and when the liquid level is higher than or equal to the incident position, n0=1 and the TIR critical angle is 60.1°. Thus, by configuring the protrusion 3120 (e.g., the angle between the first inner and outer surfaces) and / or the incident position of the light from the light source 3212, the incident angle θ i1 may be configured to be greater than 41.8° and less than 60.1°, for example, 45°. In this way, TIR may occur when the liquid level is lower than the incident position, and there is no TIR when the liquid level is higher than or the same as the incident position.

[0328] As shown in FIG. 34A, when the liquid level is lower than the incident position, the medium is gas, and TIR occurs at the first inner surface 3132, and the light is reflected within the protrusion 3120 at an incident angle θ i2 Similarly, the incident angle θ i2may be configured such that θ is greater than 41.8° and less than 60.1°, for example 45°. In this manner, TIR may occur when the liquid level is lower than the incident position, and there is no TIR when the liquid level is higher than or equal to the incident position. The angle Φ defined by the first inner surface 3132 and the second inner surface 3134 is i1 +θ i2 may be identical to

[0329] In some examples, Φ=90°, and θ i1 =θ i2 = 45°, and TIR can occur on both the first inner surface 3132 and the second inner surface 3134. out is the input light intensity I in Furthermore, the input light towards the vessel 3100 and the reflected light returning from the vessel 3100 may be parallel to each other, which can be considered a retroreflection and used to align the photodetector 3214 to receive the output light.

[0330] 34B, when the liquid level is higher than or equal to the incident position, the medium is liquid 3204, e.g., the first inner surface is in liquid 3204, no TIR occurs at the first inner surface, and light can pass through the first inner surface into liquid 3204 and further propagate outside of container 3100 with a large transmittance, e.g., about 96%. Photodetector 3214 configured to receive reflected light may not receive substantially the reflected light. In some examples, photodetector 3216 may be positioned on a different side from light source 3212 and may detect transmitted light I from container 3100. t The photodetector 3216 may be calibrated or adjusted to receive transmitted light.

[0331] In some embodiments, the detection pair 3210 includes a light source 3212 and a photodetector 3214 on the same side of the container 3100. The power of the reflected light detected by the photodetector 3214 can determine whether the liquid level is higher or lower than the incident position. For example, if the power of the detected reflected light is greater than a predetermined threshold, it indicates that TIR has occurred and the liquid level is lower than the incident position. If the power of the detected light is equal to or less than the predetermined threshold, it indicates that TIR has not occurred and the liquid level is higher than or the same as the incident position.

[0332] In some embodiments, the detection pair 3210 includes a light source 3212 and a photodetector 3216 on different sides of the container 3100. The power of the transmitted light detected by the photodetector 3216 can determine whether the liquid level is higher or lower than the incident position. For example, if the power of the detected transmitted light is less than a predetermined threshold, it indicates that TIR has occurred and the liquid level is lower than the incident position. If the power of the detected light is equal to or greater than the predetermined threshold, it indicates that TIR has not occurred and the liquid level is higher than or the same as the incident position.

[0333] FIG. 35A illustrates an exemplary liquid level measurement by moving a light source and light detector pair of a liquid level sensor along the length of a container, in accordance with one or more embodiments of the present disclosure.

[0334] The container may be the container 3100 of FIGS. 33A-33C and 34A-34B, which includes a protrusion 3120. The pair of light source and photodetector may be the pair 3210 of FIG. 34A or 34B. The pair 3210 may move continuously along the longitudinal direction 3101. The incident position of the light on the inner surface 3132 of the protrusion 3120 is at a height h relative to the bottom 3102 of the container 100. x When the pair 3210 moves, the incident position also moves, e.g., h x increases or decreases.

[0335] Corresponding height h iTo measure the liquid level 3302 in a container 3100 having a liquid level h, the coupler 3210 can be moved continuously from the bottom 3102 (h=0) towards the top 3104 of the container 3100. The intensity of the reflected light can be monitored and continuously detected. FIG. 35B shows an example measurement result 3310 using the liquid level measurement of FIG. 35A. The result 3310 shows that when the incident position h is lower than the liquid level h, TIR does not occur and the detected reflected light intensity has a lower value Ia, and when the incident position h is lower than the liquid level h i If the liquid level h is equal to or higher than Ib, TIR occurs, indicating that the intensity of the detected reflected light has a higher value Ib. i A substantial change occurs around the boundary. In some examples, the ratio of Ib / Ia may be greater than 2, 3, 4, 5, 10, 20, 50, 100, or any suitable value. In some embodiments, the predetermined threshold may be set to a value between Ia and Ib, e.g., the average of Ia and Ib, i.e., (Ia+Ib) / 2. Thus, by monitoring the detected light intensity while moving the pair 3210 along its length, the liquid level h i Accurate measurements can be obtained.

[0336] In some embodiments, a reference signal may be first detected by a photodetector, for example, detecting light reflected from a container without a protrusion. This reference signal may be used as background noise. The detected reflected light from a container with a protrusion may be subtracted from the reference signal to further improve detection accuracy.

[0337] Figure 36A shows another exemplary liquid level measurement by a liquid sensing system 3400, according to one or more embodiments of the present disclosure. Figure 36B shows an exemplary measurement result 3450 using the liquid level measurement of Figure 36A.

[0338] The liquid sensing system 3400 may include a plurality of pairs of light sources and light detectors 3210-1, 3210-2, ..., 3210-(m-1), 3210-m, 3210-(m+1), ..., 3210-(n-1), 210-n (collectively referred to as pairs 3210 or individually as pairs 3210), where m and n are integers. The pairs 3210 may be arranged at a series of positions along the longitudinal direction 3101 of the container 3100, for example, from the bottom 3102 to the top 3104. The series of positions of the pairs 3210 correspond to a series of incident positions h1, h2, ..., hm-1, hm, hm+1, ..., hn-1, hn along the longitudinal direction. The pairs 3210 may be static or attached to a support 3404 extending along the longitudinal direction. The number of pairs 3210 may be 5, 10, 20, or any suitable number.

[0339] Corresponding height h i To determine the liquid level 3402 in the container 3100 having a height h of the liquid level 3402, the respective intensities of the reflected light detected for each pair 3210 corresponding to different incident positions are obtained. The respective intensities may be plotted as individual points 3410, as shown in FIG. 36B. The incident positions, e.g., h1, h2, ... hm-1, correspond to the corresponding height h of the liquid level 3402. i If the incident positions, e.g., hm, hm+1, ..., hn-1, hn, are lower than the liquid level h, then TIR does not occur and the detected reflected light intensity has a lower intensity value Ia. i If the liquid level h is equal to or higher than h, TIR occurs and the detected reflected light intensity has a higher value Ib. i In some examples, the ratio of Ib / Ia may be greater than 2, 3, 4, 5, 10, 20, 50, 100, or any suitable value. Thus, by measuring the detected light intensity of a series of pairs 3210, the liquid level h i But, h m-1 and h m It can be determined that the value is in the range between

[0340] In some embodiments, the liquid level sensing system 400 includes one light source and light detector pair 3210 that can be positioned at a predetermined location, for example, corresponding to a predetermined liquid level or a predetermined volume level. Liquid can be gradually poured into the container 3100. The light detector can continuously monitor the reflected light from the container 3100. If a substantial change occurs, indicating that TIR has occurred and the liquid level has increased to the predetermined liquid level, the liquid level sensing system 3400 can send a signal to stop pouring the liquid. In this way, a precise amount of liquid can be obtained in the container 3100. In some embodiments, the liquid level sensing system 3400 includes two or more pairs 3210 to control the pouring of a fixed amount of liquid into the container.

[0341] 37 shows an exemplary fluid system 3500 for liquid level sensing, in accordance with one or more embodiments of the present disclosure. The fluid system 3500 can include at least one input container 3502, a fluid channel system 3504 including one or more fluid channels, and one or more output containers 3506.

[0342] In some embodiments, the liquid (e.g., a mixture of solutions) contained in the input container 3502 can be flowed to the fluid channel system 3504 for processing, e.g., filtration, separation, or any suitable processing step. The fluid channel system 3504 can direct the processed liquid to one or more output containers 3506. In some embodiments, the one or more output containers 3506 include at least one sample container 3506 and at least one waste container 3506.

[0343] In some embodiments, at least one of the at least one input container 3502 and the one or more output containers 3506 may be container 3100 and may be assembled with one or more pairs of light source and light detector 3210 to form one or more liquid level sensors 3200 or 3300, or system 3400. In this manner, the initial and / or final volume of liquid in input container 3502 may be determined. The volume of sample or waste in output container 3506 may be determined. In some cases, the difference between the volume of liquid in the input container and the volume of sample or waste in output container 3506 may be obtained for further analysis.

[0344] FIG. 38 is a flowchart of an exemplary process 3600 for liquid level sensing using optical reflection, according to one or more embodiments of the present disclosure. Process 3600 may be performed by a liquid level sensor (e.g., liquid level sensor 3200 of FIGS. 34A-34B, liquid level sensor 3300 of FIG. 35A), a liquid level system (e.g., liquid level system 3400 of FIG. 36A), or a fluid system (e.g., fluid system 3500 of FIG. 37). Liquid is held in a container, such as container 3100 of FIGS. 33A-33C, 34A-34B, 34A, 36A, or 36. The liquid level sensor includes a pair of a light source (e.g., 3212 of FIGS. 34A-34B) and a light detector (e.g., 3214 or 3216 of FIGS. 34A-34B).

[0345] Light is emitted 3602, for example from a light source, onto an exterior surface of a container containing a liquid, the container including at least one interior surface configured such that i) the light is totally reflected internally by the at least one interior surface when the level of the liquid in the container is lower than the incident position where the light is incident on the at least one interior surface, and ii) the light is transmitted through the at least one interior surface when the level of the liquid is higher than or equal to the incident position where the light is incident on the at least one interior surface.

[0346] Light (eg, transmitted or reflected light) from the container is detected (3604), for example, by a photodetector, and the level of liquid in the container is then determined (3606) based on the detected light.

[0347] The container can extend along a longitudinal direction (e.g., 3101 in FIG. 33A ), and the liquid level can be defined from the bottom of the container along the longitudinal direction. In some embodiments, the container includes a protrusion (e.g., 3120 in FIGS. 33A-33C , 34A-34B , 35A , or 36A ) having at least one inner surface. The protrusion can extend continuously along the longitudinal direction or can include multiple portions spaced apart along the longitudinal direction.

[0348] In some embodiments, the protrusion includes a first inner surface (e.g., 3132 in FIGS. 33B-33C or 34A-34B) on a first side (e.g., 3122 in FIGS. 33B-33C) and a second inner surface (e.g., 3134 in FIGS. 33B-33C or 34A-34B) on a second side (e.g., 3124 in FIGS. 33B-33C). As shown in FIG. 34A, light may be totally internally reflected by the first inner surface and then totally internally reflected by the second inner surface when the liquid level in the container is at or above the incident position (e.g., higher than the incident position along the longitudinal direction).

[0349] In some embodiments, the angle defined by the first and second inner surfaces is substantially equal to 90 degrees, as shown, for example, in Figures 33C and 34A-34B. The first angle of incidence at which light is incident on the first inner surface can be substantially equal to the second angle of incidence at which light is incident on the second inner surface. For example, each of the first and second angles of incidence can be substantially equal to 45 degrees.

[0350] In some embodiments, the first and second sides of the protrusion are connected together at an edge, e.g., 3123 in FIG. 33B. The container may include a body (e.g., 3110 in FIGS. 33A-33C or 34A-34B) for containing a liquid. The protrusion may include a third side (e.g., 3126 in FIG. 33B) that is part of the body of the container, and the first and second sides are externally connected to the body of the container. In some examples, the protrusion has a triangular shape with a first side, a second side, and a third side, as shown, for example, in FIG. 33C or 34A-34B.

[0351] In some embodiments, the body comprises a first material and the protrusion comprises a second material. In some cases, the second material is the same as the first material, and the protrusion and body are integral. In some cases, the second material is different from the first material, and the protrusion and body are attached together.

[0352] In some embodiments, light is incident perpendicularly on the outer surface of the container and exits perpendicularly from the container, for example as shown in Figures 34A-34B. In some embodiments, when the liquid level in the container is lower than the point of incidence, as shown in Figure 34A, the light propagates along a first direction, incident on a first inner surface, and is reflected from a second inner surface along a second direction that is substantially parallel to but opposite the first direction.

[0353] Light is emitted from a light source, and the light is detected by a photodetector, and the light source and the photodetector form a pair. The incident position can be predetermined based on the position of the light source, for example, as shown in Figure 35A or 36A.

[0354] In some embodiments, the light source and the light detector are positioned on the same side of the container, for example, as in Figures 34A and 34B. Determining the level of liquid in the container based on the detected light may include determining that the level of liquid is lower than the incident position in response to determining that the power of the detected light is greater than a predetermined threshold, or determining that the level of liquid is higher than or the same as the incident position in response to determining that the power of the detected light is equal to or less than the predetermined threshold.

[0355] In some embodiments, the light source and the light detector are positioned on opposite sides of the container, for example, as shown in Figure 34B. Determining the level of liquid in the container based on the detected light may include determining that the level of liquid is lower than the incident position in response to determining that the power of the detected light is less than a predetermined threshold, or determining that the level of liquid is higher than or the same as the incident position in response to determining that the power of the detected light is equal to or greater than the predetermined threshold.

[0356] In some embodiments, as shown in Figures 35A-35B, process 3600 may include gradually moving the container or at least one of the pair of light source and light detector along a longitudinal direction until the detected light power changes substantially beyond a predetermined threshold, and determining the level of the liquid based on the position of the bottom of the container and the position of the light source when the detected light power changes substantially beyond the predetermined threshold.

[0357] In some embodiments, process 3600 may include monitoring the detected power of the light while the level of the liquid in the container is increased by injecting the liquid into the container, where the position of incidence of the light corresponds to a predetermined level in the container, and controlling to stop the injection in response to determining that the detected power of the light has changed substantially beyond a predetermined threshold.

[0358] In some embodiments, light is emitted from a plurality of longitudinally spaced light sources, and the light is detected by a plurality of longitudinally spaced light detectors, as shown in Figure 36A. Each light source of the plurality of light sources can be associated with a respective light detector of the plurality of light detectors and configured to emit a corresponding portion of light, and each light detector can be configured to detect a corresponding portion of light emitted from the light source.

[0359] In some cases, each light source and each photodetector of the multiple light sources are arranged in a plane passing through the longitudinal axis of the container. For example, if the beam size of the light is relatively small compared to the size of the container, the light is emitted into the container, reflected within the protrusion, and reflected out of the container in the same plane, and each photodetector can be arranged side by side with the light source in the same plane. The light emitted from the light source can be configured to be focused at least in the longitudinal direction so that the incident positions on the first and second inner surfaces are both in a plane perpendicular to the longitudinal direction. The distance between the light source and each photodetector can be determined based on the light input and output positions of the container. As shown in FIG. 34A, the farther the input position is from the intersection of the two inner surfaces (e.g., edge 3133 in FIG. 33C) along the longitudinal direction, the greater the distance. That is, after the distance between the light source and each photodetector is determined, the light input position of the body can be adjusted, for example, to maximize the intensity of the reflected light detected by each photodetector.

[0360] In some cases, each light source and each photodetector of the multiple light sources are arranged along the longitudinal direction. For example, if the beam size of the light is relatively large compared to the size of the container, the light is emitted into the container, reflected within the protrusion, and reflected out of the container at different planes along the longitudinal direction. Each photodetector can be arranged below or above the light source along the horizontal direction. The position of each photodetector can be calibrated or adjusted to maximize the intensity of the reflected light detected.

[0361] The protrusion extends vertically, and multiple depths (volumes) are measured at the protrusion. The light source and photodetector can be arranged either side by side or one above the other, provided they are within the width of the protrusion. Specular reflection can double the molding draft. If the molding draft is 1 degree, the protrusion can reflect the incident light downward by 2 degrees. In some embodiments, the photodetector is placed below the light source, for example, to account for alignment tolerances. The distance between the incident and returning light beams cannot be greater than the width of the protrusion. With the photodetector below the light source, the pair of photodetector and light source can move horizontally along almost the entire width of the protrusion, since the first and second reflections on the first and second inner surfaces can occur on either surface.

[0362] In some embodiments, determining the level of liquid in the container based on the detected light includes determining the level of liquid in the container based on corresponding portions of light detected by each of the plurality of photodetectors. For example, as shown in FIGS. 36A-36B, step 3606 may include: determining, based on a first corresponding portion of light detected by a first photodetector, that the level of liquid in the container is higher than a first incident position where a first corresponding portion of light from a first light source is incident on a first inner surface; determining, based on a second corresponding portion of light detected by a second photodetector, that the level of liquid in the container is lower than a second incident position where a second corresponding portion of light from a second light source is incident on the first inner surface; and determining that the level of liquid in the container is between the first incident position and the second incident position. The first incident position is predetermined based on a first position of the first light source, and the second incident position is predetermined based on a second position of the second light source.

[0363] In some embodiments, the container is a tube, vessel, tank, bottle, or any suitable structure configured to hold a liquid.

[0364] In some embodiments, the container may be formed by forming the body of the container and forming the protrusions together, such that the body and protrusions are formed as a single unit, for example, by SLA 3D printing or molding, hi some embodiments, the container is formed by attaching the protrusions to the body of the container.

[0365] In some embodiments, the liquid level sensor includes at least one processor configured to determine the level of the liquid based on the light detected by the light detector. In some embodiments, the liquid level sensor is coupled to the at least one processor and configured to provide the detected light to the at least one processor configured to determine the level of the liquid based on the detected light. In some embodiments, the at least one processor is configured to multiplex several liquid level sensors, for example, by controlling one liquid level sensor to be on at a time to prevent crosstalk due to light piping and scattering.

[0366] In some embodiments, as shown in Figure 37, a fluid system (e.g., 3500 in Figure 37) includes at least one input container configured to contain a liquid, at least one fluid channel configured to receive liquid from the input container and allow the liquid to flow out of at least one outlet, and at least one effluent container coupled to the at least one outlet and configured to receive liquid through the at least one fluid channel, At least one of the input container or the at least one effluent container is configured for liquid level detection, each of the at least one container includes at least one inner surface configured such that i) when the liquid level in the container is lower than an incident position where the light is incident on the at least one inner surface, light is totally reflected internally by the at least one inner surface, and ii) when the liquid level is higher than or equal to an incident position where the light is incident on the at least one inner surface, light is transmitted through the at least one inner surface.

[0367] The fluid system may further include at least one processor configured to determine a level of liquid in the container based on the detected light from the container, as described above. In some embodiments, the at least one processor is configured to determine a difference between a volume of liquid in the input container and a volume of liquid in the at least one output container.

[0368] MicroOrganoSphere (MOS)

[0369] Historically, the most common approach for testing cellular responses to various stimuli involves culturing cells in two-dimensional (2D) conditions, such as in Petri dishes or well plates. However, these 2D conditions can be stressful for cells and do not necessarily correlate well with individual patient responses to the tested stimuli. Liquid droplets, on the other hand, can provide a three-dimensional (3D) environment for clusters of 3D cell aggregates, potentially resulting in experimental results that correlate better with patient outcomes. 3D cell aggregates can include organoids or spheroids. Organoids are in vitro cell aggregates, typically greater than 1 mm in diameter, containing populations of stem cells capable of differentiating into cells of major cell lineages. Spheroids are simple clusters of cells from a wide range of sources, such as tumor tissue, embryoid bodies, hepatocytes, neural tissue, or mammary glands. Unlike organoids, spheroids typically lack the ability to self-organize or regenerate.

[0370] The droplets generated to contain 3D cellular aggregates, such as organoids or spheroids, are sometimes referred to as "Patient-Derived MicroOrganoSpheres" (PMOS) or simply "MicroOrganoSpheres" (MOS). MOS, including methods and apparatus for generating them, are described in further detail in U.S. Patent Application No. 16 / 838,010, which is incorporated herein by reference in its entirety.

[0371] It is important to note that while MOS is described herein as one example of droplets that can be generated by microfluidic devices, this disclosure is not intended to be limiting. Microfluidic-formed droplets can be used in a wide range of applications, including drug discovery and drug efficacy evaluation, and those skilled in the art will understand that the techniques described herein can be applied to droplets used in many of these alternative applications.

[0372] In some examples, MOSs contain dissociated cells from a patient-derived tissue sample (e.g., a biopsy sample). The tissue can be derived from a healthy tissue biopsy or a cancerous (e.g., tumor) cell biopsy. The cells are dissociated and dispersed (e.g., suspended) in a matrix material. Once generated, the MOSs can be patterned onto a microfluidic microwell array, incubated, administered with a drug compound, and imaged at repeated time intervals to monitor the growth of each organoid. This miniaturized assay maximizes the use of patient samples and enables high-throughput screening of numerous therapeutic agents (e.g., drugs or drug formulations) from a core biopsy at a relatively low cost per sample.

[0373] The matrix material used to generate MOS can be a liquid, such as a gel, semi-solid, or low-viscosity liquid, at room temperature (e.g., about 25°C). Exemplary matrix materials include polymers or hydrogels including collagen, fibrin, or chitosan; MATRIGEL™ (Corning, Corning, NY); polyethylene glycol; dextran, including chemically or photocrosslinkable dextran; electrospun biological, synthetic, or biological-synthetic blends; or other suitable polymerizable matrix materials. In some examples, the matrix material is a gel, such as a synthetic gel or a natural gel. Examples of synthetic gels include gels derived from polyethylene glycol (PEG), polyhydroxyethyl methacrylate (PHEMA), polyvinyl alcohol (PVA), or polyethylene oxide (PEO).

[0374] In some instances, upon polymerization, the matrix material forms a hydrogel. The term "hydrogel" refers to a two- or multi-component gel comprising a three-dimensional network of polymer chains, with water acting as the dispersion medium and filling the spaces between the polymer chains. Exemplary hydrogels that can be used in MOS include alginate, collagen (including type I and type VI collagen), elastin, keratin, fibronectin, proteoglycans, glycoproteins, polylactides, polyethylene glycols, polycaprolactones, polycolides, polydioxanones, polyacrylates, polyurethanes, polysulfones, peptide sequences, proteins and derivatives, oligopeptides, gelatin, elastin, fibrin, laminin, polymethacrylates, polyacetates, polyesters, polyamides, polycarbonates, polyanhydrides, polyamino acids, carbohydrates, polysaccharides and modified polysaccharides, or derivatives and copolymers thereof; glasses, such as bioactive glasses; ceramics; silica; alumina; calcite; hydroxyapatite; calcium phosphate; inorganic materials, such as bone; or combinations of the foregoing. In certain examples, the hydrogel comprises a material selected from the group consisting of agarose, alginate, type I collagen, polyoxyethylene-polyoxypropylene block copolymer (e.g., Pluronic® F127 (BASF Corporation, Mount Olive, NJ)), silicone, polysaccharide, polyethylene glycol, and polyurethane.

[0375] In some examples, the MOS includes one or more biologically relevant materials in addition to the patient-derived cells and matrix material. Exemplary biologically relevant materials included in the MOS may include one or more of the following: extracellular matrix proteins (e.g., fibronectin), drugs (e.g., small molecules), peptides, antibodies (e.g., for modulating either cell survival, proliferation, or differentiation), or inhibitors of specific cellular functions. Biologically relevant materials in the MOS may be used, for example, to increase cell viability by reducing cell death and / or activating cell growth / replication, or to otherwise mimic an in vivo environment. Biologically relevant materials incorporated into the MOS may include or mimic one or more of the following components: serum, interleukins, chemokines, growth factors, glucose, physiological salts, amino acids, or hormones. When the matrix material is a gel, the gel itself may include one or more biologically relevant materials, including extracellular matrix components such as collagen, fibrinogen, laminin, fibronectin, vitronectin, hyaluronic acid, fibrin, alginate, agarose, or chitosan. For example, MATRIGEL™ contains bioactive polymers that are important for cell viability, proliferation, development, and migration. In a particular example, the matrix material is a gel that includes type I collagen, such as type I collagen obtained from rat tail, either alone or in addition to other biologically relevant materials, such as other extracellular matrix proteins.

[0376] The MOS described herein may have a diameter of about 50 μm to about 500 μm (e.g., about 50 μm to about 400 μm, about 50 μm to about 300 μm, about 50 μm to about 250 μm, etc.). Each MOS may initially contain about 1 to 1,000 dissociated primary cells (e.g., about 1 to 750, about 1 to 500, about 1 to 400, about 1 to 300, about 1 to 200, about 1 to 150, about 1 to 100, about 1 to 75, about 1 to 50, about 1 to 40, about 1 to 30, about 1 to 20, etc.) distributed within the matrix material. The number of cells per MOS may be determined based on the intended use of the MOS. For example, MOSs with a small number of cells per MOS (e.g., 1-5 cells per MOS) may be useful for studying clonal diversity (e.g., tumor heterogeneity), e.g., observing which clones are drug-resistant and determining the genomic (mutational) diversity associated with specific clones (e.g., by genome sequencing). MOSs with a moderate number of cells per MOS (e.g., approximately 3-30 cells, 5-30 cells, 5-25 cells, 5-20 cells, 10-25 cells, etc.) may be useful for rapid drug testing, e.g., toxicity testing, because these MOSs tend to grow rapidly. Because MicroOrganoSpheres may contain different lineages, potentially including epithelial (e.g., cancer) and mesenchymal (e.g., stromal, immune, vascular) cells, MOSs with a large number of cells per MOS (e.g., approximately 20-100 cells, e.g., 30-100 cells, 40-100 cells, or more than 50 cells) may be suitable for mimicking tissue composition in each MicroOrganoSphere.

[0377] MOS generated by the microfluidic systems described herein can be used substantially immediately after formation or can be cultured for short periods of time (e.g., 14 days or less, 10 days or less, 7 days or less, 5 days or less, etc.). Cells within MOS can survive while maintaining many, but not all, characteristics of the tissue, including tumor tissue, from which they were extracted. Cell survival within MOS is high, and MOS can be cultured for periods of time (e.g., days or weeks) through multiple passages, during which the cells divide, cluster, and form structures similar to the parent tissue, sometimes referred to as tumor-like masses. In some cases, cells from dissociated tissue within the MOS form morphological structures within the MOS.

[0378] 39A-39C show exemplary MOSs. Specifically, referring to FIG. 39A, after generation, each MOS contains a single cell and has a diameter of approximately 300 μm. FIG. 39B shows MOSs after three days in culture. The cells have expanded in size, e.g., doubling and / or growing. Referring to FIG. 39C, after seven days in culture, the cells have doubled multiple times and formed clusters or clumps of cells, or tumor-like masses.

[0379] 40 illustrates an example of a computing device 2800 and a mobile computing device 2850 that may be used to implement embodiments of the present disclosure. The computing device 2800 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The mobile computing device 2850 is intended to represent various forms of mobile devices, such as personal digital assistants, mobile phones, smartphones, augmented reality devices, sensor devices, smart cameras, and other similar computing devices. The components, their connections and relationships, and their functions shown here are intended to be exemplary only and not limiting. The computing device 2800 and / or the mobile computing device 2850 may form at least a portion of a microfluidic system, such as the controller 2124 of the device 2100 described above. Also, computing device 2800 and / or mobile computing device 2850 may form at least a portion of a remote computing device external to device 2100, which may perform one or more of the image processing operations described in connection with Figures 28, 29, and 30, or which may interact with device 2100 to receive data transmitted from device 2100 (e.g., a captured image and / or an estimated size of one or more droplets). For example, in some implementations, computing device 2800 and / or mobile computing device 2850 may form at least a portion of a remote camera or smart camera external to device 2100.

[0380] Computing device 2800 includes a processor 2802 (e.g., a digital signal processor [DSP], a graphics processing unit [GPU], a field programmable gate array [FPGA], etc.), a memory 2804, a storage device 2806, a high-speed interface 2808, and a low-speed interface 2812. In some implementations, the high-speed interface 2808 is connected to the memory 2804 and a plurality of high-speed expansion ports 2810. In some implementations, the low-speed interface 2812 is connected to a low-speed expansion port 2814 and the storage device 2804. Each of the processor 2802, memory 2804, storage device 2806, high-speed interface 2808, high-speed expansion port 2810, and low-speed interface 2812 are interconnected using various buses and may be mounted on a common motherboard or in other manners as desired. Processor 2802 can process instructions for execution within computing device 2800, including instructions stored in memory 2804 and / or storage device 2806, to display graphical information for a graphical user interface (GUI) on an external input / output device, such as a display 2816 coupled to high-speed interface 2808. In other implementations, multiple processors and / or multiple buses can be used, along with multiple memories and types of memory, as appropriate. Additionally, multiple computing devices can be connected, each providing a portion of the required operations (e.g., as a bank of servers, a group of blade servers, or a multiprocessor system).

[0381] The memory 2804 stores information within the remote device 2800. In some implementations, the memory 2804 is a volatile memory unit(s). In some implementations, the memory 2804 is a non-volatile memory unit(s). The memory 2804 may also be another form of computer-readable medium, such as a magnetic disk or an optical disk.

[0382] The storage device 2806 can provide mass storage for the computing device 2800. In some implementations, the storage device 2806 can be or include a computer-readable medium such as a floppy disk device, a hard disk device, an optical disk device, a tape device, a flash memory or other similar solid-state memory device, or an array of devices, including devices in a storage area network or other configuration. The instructions can be stored on an information carrier. When executed by one or more processing devices, such as the processor 2802, the instructions perform one or more methods, such as those described above. The instructions can also be stored by one or more storage devices, such as a computer-readable or machine-readable medium, such as the memory 2804, the storage device 2806, or a memory on the processor 2802.

[0383] High-speed interface 2808 manages bandwidth-intensive operations for computing device 2800, while low-speed interface 2812 manages low-bandwidth-intensive operations. Such an allocation of functionality is merely exemplary. In some implementations, high-speed interface 2808 is coupled to memory 2804, display 2816 (e.g., via a graphics processor or accelerator), and high-speed expansion port 2810, which can accept various expansion cards. In some implementations, low-speed interface 2812 is coupled to storage device 2806 and low-speed expansion port 2814. Low-speed expansion port 2814, which may include various communication ports (e.g., Universal Serial Bus (USB), Bluetooth, Ethernet, Wireless Ethernet), may be coupled to one or more input / output devices. Such input / output devices may include a display device, a printing device 2834, or a keyboard or mouse 2836. Input / output devices may also be coupled to low-speed expansion port 2814 via a network adapter. Such network input / output devices may include, for example, a switch or router 2832 .

[0384] Computing device 2800 may be implemented in several different forms, as shown in FIG. 40 . For example, computing device 2800 may be implemented as a standard server 820 or multiple times within a group of such servers. Additionally, computing device 2800 may be implemented in a personal computer, such as a laptop computer 2822. Also, computing device 2800 may be implemented as part of a rack server system 2824. Alternatively, components of computing device 2800 may be combined with other components in a mobile device, such as mobile computing device 2850. Each such device may house one or more of computing device 2800 and mobile computing device 2850, and the overall system may be made up of multiple computing devices communicating with each other.

[0385] The mobile computing device 2850 includes, among other components, a processor 2852, memory 2864, an input / output device such as a display 2854, a communication interface 2866, and a transceiver 2868. The mobile computing device 2850 may also be provided with a storage device such as a microSD card or other device to provide additional storage. Each of the processor 2852, memory 2864, display 2854, communication interface 2866, and transceiver 2868 are interconnected using various buses, and some of the components may be mounted on a common motherboard or in other manners as desired. In some implementations, the mobile computing device 2850 may include camera device(s).

[0386] The processor 2852 can execute instructions within the mobile computing device 2850, including instructions stored in the memory 2864. The processor 2852 can be implemented as a chipset of chips including discrete and multiple analog and digital processors. For example, the processor 2852 can be a complex instruction set computer (CISC) processor, a reduced instruction set computer (RISC) processor, or a minimal instruction set computer (MISC) processor. The processor 2852 can provide coordination of other components of the mobile computing device 2850, such as control of a user interface (UI), applications executed by the mobile computing device 2850, and / or wireless communication by the mobile computing device 2850.

[0387] The processor 2852 may communicate with a user via a control interface 2858 and a display interface 2856 coupled to a display 2854. The display 2854 may be, for example, a thin film transistor liquid crystal display (TFT) display, an organic light emitting diode (OLED) display, or other suitable display technology. The display interface 2856 may include appropriate circuitry for driving the display 2854 to present graphics and other information to the user. The control interface 2858 may receive commands from the user and convert them for transmission to the processor 2852. Additionally, an external interface 2862 may provide communication with the processor 2852 to enable short-range communication of the mobile computing device 2850 with other devices. The external interface 2862 may provide, for example, wired communication in some implementations or wireless communication in other implementations, and multiple interfaces may also be used.

[0388] The memory 2864 stores information within the mobile computing device 2850. The memory 2864 may be implemented as one or more of a computer-readable medium(s), a volatile memory unit(s), or a non-volatile memory unit(s). Expansion memory 2874 may also be provided and connected to the mobile computing device 2850 via an expansion interface 2872, which may include, for example, a single in-line memory module (SIMM) card interface. The expansion memory 2874 may provide additional storage space for the mobile computing device 2850 or may also store applications or other information for the mobile computing device 2850. In particular, the expansion memory 2874 may include instructions that perform or complement the processes described above and may also include secure information. Thus, for example, the expansion memory 2874 may be provided as a security module for the mobile computing device 2850 and may be programmed with instructions that enable secure use of the mobile computing device 2850. Additionally, secure applications may be provided via SIMM cards along with additional information such as placing identifying information on the SIMM card in an unhackable manner.

[0389] The memory may include, for example, flash memory and / or non-volatile random access memory (NVRAM), as discussed below. In some embodiments, the instructions are stored on an information carrier. The instructions, when executed by one or more processing devices, such as processor 2852, perform one or more methods, such as those described above. The instructions may also be stored by one or more storage devices, such as one or more computer-readable or machine-readable media, such as memory 2864, expansion memory 2874, or memory on processor 2852. In some embodiments, the instructions may be received in a propagated signal, such as via transceiver 2868 or external interface 2862.

[0390] The mobile computing device 2850 may communicate wirelessly via a communications interface 2866, which may include digital signal processing circuitry as needed. Communications interface 2866 may provide communications under various modes or protocols, such as Global System for Mobile Communications (GSM) voice calls, Short Message Service (SMS), Enhanced Messaging Service (EMS), Multimedia Messaging Service (MMS) messaging, Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Personal Digital Cellular (PDC), Wideband Code Division Multiple Access (WCDMA), CDMA2000, General Packet Radio Service (GPRS), and the like. Such communications may occur via a transceiver 2868, for example, using radio frequencies. Additionally, short-range communications may also occur, such as using Bluetooth or Wi-Fi. Additionally, a Global Positioning System (GPS) receiver module 2870 may provide additional navigation-related and location-related radio data to the mobile computing device 2850, which may be used as needed by applications executing on the mobile computing device 2850.

[0391] The mobile computing device 2850 may also communicate audibly using an audio codec 2860, which may receive spoken information from a user and convert it into usable digital information. The audio codec 2860 may also generate audible sounds for the user, for example, via a speaker in the handset of the mobile computing device 2850. Such sounds may include sounds from a voice telephone call, may include recorded sounds (e.g., voice messages, music files, etc.), and may include sounds generated by applications running on the mobile computing device 2850.

[0392] The mobile computing device 2850 may be implemented in several different forms, as shown in Figure 40. For example, the mobile computing device 2850 may implement a telephone device 2880, a personal digital assistant 2882, and a tablet device (not shown). The mobile computing device 2850 may also be implemented as a component of a smartphone, an AR device, or other similar mobile device.

[0393] Computing device 2800 may be implemented as part of a microfluidic system, such as controller 2124 of device 2100 described above with respect to Figure 26A. Computing device 2800 may also be implemented in a remote computing device that communicates with device 2100 and receives data transmitted from device 2100 (e.g., captured images and estimated sizes of one or more droplets).

[0394] Computing device 2800 and / or 2850 may also include a USB flash drive, which may store an operating system and other applications, and which may include input / output components such as a wireless transmitter or a USB connector that can be inserted into a USB port on another computing device.

[0395] Enumeration of Embodiments Embodiment 1. A microfluidic device comprising: A microfluidic chip for generating MicroOrganoSpheres (MOS), the microfluidic chip having a first microfluidic channel defined on a surface thereof, the first microfluidic channel comprising: a droplet generation portion including an inlet portion, a junction between the inlet portion and the emulsification fluid channel, and a chamber downstream of the junction, the cross-sectional area of ​​the chamber being greater than the cross-sectional area of ​​the inlet portion; a microfluidic chip including a polymerization section downstream of the droplet generation section, the polymerization section having a serpentine shape; A cartridge for MOS demulsification, comprising: collection containers, a substrate disposed over the collection reservoir, the substrate having a second microfluidic channel defined in a surface of the substrate facing the collection reservoir, the second microfluidic channel fluidly connected to an outlet of the polymerization portion of the first microfluidic channel; and a cartridge comprising a membrane disposed between the collection reservoir and the surface of the substrate.

[0396] Embodiment 2. A microfluidic device as described in embodiment 1, wherein the droplet generation portion of the first microfluidic channel comprises an outlet portion downstream of a chamber, and the cross-sectional area of ​​the chamber is larger than the cross-sectional area of ​​the outlet portion.

[0397] Embodiment 3. The microfluidic device of embodiment 2, wherein at least some of the outlet portions extend in a direction parallel to the chamber.

[0398] Embodiment 4. The microfluidic device of any of the preceding embodiments, wherein the surface of the microfluidic chip is a first surface, and wherein the polymerized portion of the microfluidic channel is defined on the first surface of the microfluidic chip and on a second surface of the microfluidic chip opposite the first surface.

[0399] Embodiment 5. The microfluidic device of any preceding embodiment, wherein the junction comprises a junction with two hydrophobic fluidic channels.

[0400] Embodiment 6. The microfluidic device of embodiment 5, wherein the junction is a right-angle junction.

[0401] Embodiment 7. The microfluidic device of any preceding embodiment, wherein the membrane comprises a hydrophobic membrane.

[0402] Embodiment 8. A microfluidic device according to any of the preceding embodiments, wherein the second microfluidic channel comprises an upstream section having a simple serpentine shape and a downstream section having a double serpentine shape.

[0403] Embodiment 9. A microfluidic device according to any preceding embodiment, wherein the cross-sectional area of ​​the second microfluidic channel decreases from the inlet end of the second microfluidic channel to the outlet end of the second microfluidic channel.

[0404] Embodiment 10. A microfluidic device according to any of the preceding embodiments, wherein the surface of the substrate is a first surface, and wherein a medium inlet channel is defined on a second surface of the substrate opposite the first surface of the substrate, the medium inlet channel being fluidly connected to an upstream section of the second microfluidic channel and configured to be connected to a medium reservoir.

[0405] Embodiment 11. The microfluidic device of embodiment 10, wherein the demulsifying cartridge comprises a medium reservoir.

[0406] Embodiment 12. A microfluidic device according to embodiment 11, wherein the media inlet channel is fluidly connected to the media reservoir via a tube extending through the substrate and the collection vessel.

[0407] Embodiment 13. A microfluidic device according to embodiment 11 or 12, wherein the collection vessel is disposed within a cavity defined in the media reservoir such that the collection vessel is positioned between the media reservoir and the substrate.

[0408] Embodiment 14. A microfluidic device according to any one of embodiments 11 to 13, wherein the bottom surface of the medium reservoir is angled relative to the plane of the substrate.

[0409] Embodiment 15. A microfluidic device described in any of embodiments 11 to 14, wherein the demulsification cartridge includes a duckbill valve extending through the substrate and the collection vessel, the duckbill valve configured to provide fluid access to the media reservoir.

[0410] Embodiment 16. The microfluidic device of any preceding embodiment, wherein the demulsifying cartridge comprises a hydrophobic material disposed within the collection reservoir.

[0411] Embodiment 17. The microfluidic device of any preceding embodiment, wherein a vacuum channel is defined through the body of the collection reservoir, the vacuum channel being configured to allow a vacuum to be applied to a surface of the membrane opposite the substrate.

[0412] Embodiment 18. A microfluidic device according to any of the preceding embodiments, comprising a reservoir fluidly connected to the first microfluidic channel via an inlet port defined at the inlet end of the first microfluidic channel.

[0413] Embodiment 19. A microfluidic device according to embodiment 18, wherein the reservoir comprises a base and a cover, the base and the cover defining a cavity for the fluid sample.

[0414] Embodiment 20. The microfluidic device of embodiment 19, comprising an inlet port in the cover of the reservoir, the inlet port comprising a duckbill valve.

[0415] Embodiment 21. A microfluidic device according to embodiment 19 or 20, comprising an outlet port in the cover of the reservoir, the outlet port being connected to a tube extending into the cavity of the reservoir.

[0416] Embodiment 22. A microfluidic device according to any one of embodiments 19 to 21, wherein the bottom surface of the base of the reservoir is angled relative to the cover.

[0417] Embodiment 23. A microfluidic device according to any of embodiments 18 to 22, comprising a reservoir holder configured to receive a reservoir, the reservoir holder comprising a cooling system configured to cool the reservoir.

[0418] Embodiment 24. A microfluidic device as described in embodiment 23, wherein the cooling system comprises a thermoelectric cooling system.

[0419] Embodiment 25. A microfluidic device according to any of the preceding embodiments, wherein one or more cutouts are defined between the droplet generation portion and the polymerization portion of the microfluidic chip.

[0420] Embodiment 26. A microfluidic device as described in embodiment 25, wherein the edges of one or more cutouts are angled relative to the surface of the microfluidic chip.

[0421] Embodiment 27. A microfluidic device according to embodiment 25 or 26, wherein one or more cutouts extend through the entire thickness of the microfluidic chip.

[0422] Embodiment 28. A microfluidic device according to any preceding embodiment, comprising a cover disposed on the surface of the microfluidic chip.

[0423] Embodiment 29. The microfluidic device of embodiment 28, wherein the cover comprises an optically transparent cover.

[0424] Embodiment 30. A microfluidic device according to any of the preceding embodiments, wherein a plurality of first microfluidic channels are defined on the surface of the microfluidic chip, and wherein the device includes a plurality of cartridges, the second microfluidic channels of each cartridge being fluidically connected to a corresponding one of the first microfluidic channels of the microfluidic chip.

[0425] Embodiment 31. A microfluidic device according to any of the preceding embodiments, wherein the device includes an effluent vial fluidly connected to the second microfluidic channel via an outlet port defined in the outlet end of the second microfluidic channel.

[0426] Embodiment 32. A system that can be combined with any of embodiments 1 to 31, comprising a microfluidic device according to any of the preceding embodiments, a housing in which the microfluidic device is disposed, and a polymerization block housed within the housing and positioned to apply a stimulus to a polymerization portion of the first microfluidic channel.

[0427] Embodiment 33. The system of embodiment 32, wherein the polymerization block comprises a thermal polymerization block configured to apply heat to the polymerization portion of the first microfluidic channel.

[0428] Embodiment 34. The system of embodiment 33, wherein the thermal polymerization block comprises a heater.

[0429] Embodiment 35. The system of embodiment 34, wherein the thermal polymerization block includes a temperature sensor.

[0430] Embodiment 36. A system as described in embodiment 35, wherein the temperature sensor includes one or more of a thermistor, a thermocouple, or a resistance temperature detector.

[0431] Embodiment 37. A system described in embodiment 35 or 36, including a controller configured to control operation of the resistive heater in response to temperature data received from the temperature sensor.

[0432] Embodiment 38. A system described in any of embodiments 34 to 37, wherein the heater comprises a resistive heater.

[0433] Embodiment 39. A system described in any of embodiments 34 to 38, wherein the thermal polymerization block includes an insulating cover and the heater is disposed within a cavity defined within the insulating cover.

[0434] Embodiment 40. A system described in any of embodiments 32 to 39, wherein the polymerization block includes a photopolymerization block configured to illuminate the polymerization portion of the first microfluidic channel.

[0435] Embodiment 41. The system of embodiment 40, wherein the photopolymerization block comprises a light-emitting diode (LED).

[0436] Embodiment 42. The system of embodiment 41, wherein the photopolymerization block comprises a photodetector.

[0437] Embodiment 43. The system of embodiment 42, including a controller configured to control operation of the LED in response to light intensity data received from the photodetector.

[0438] Embodiment 44. A system described in any of embodiments 41 to 43, wherein the LED is disposed within a cavity defined within the housing of the photopolymerization block.

[0439] Embodiment 45. A system described in embodiment 44, wherein the walls of the cavity are formed from a material capable of reflecting light at the wavelength of the light output by the LED.

[0440] Embodiment 46. A system described in any of embodiments 41 to 45, including a controller configured to control the LED to emit pulsed illumination.

[0441] Embodiment 47. A system described in any of embodiments 32 to 46, wherein the surface of the microfluidic chip is a first surface, and the polymerized block includes a first block positioned adjacent to the first surface of the microfluidic chip and a second block positioned adjacent to the second surface of the microfluidic chip, and the second surface is opposite the first surface.

[0442] Embodiment 48. A system described in embodiment 47, wherein the first block and the second block are fixed relative to the microfluidic chip by one or more springs.

[0443] Embodiment 49. A system described in embodiment 47 or 48, wherein the first block and the second block are clamped to a microfluidic chip.

[0444] Embodiment 50. A system described in any of embodiments 32 to 49, comprising a reservoir for an emulsifying fluid, the emulsifying fluid channel of the microfluidic device being fluidically connected to the reservoir.

[0445] Embodiment 51. A system as described in embodiment 50, wherein the reservoir includes a reflective rib for fluid volume measurement disposed within the chamber of the reservoir.

[0446] Embodiment 52. A system described in embodiment 50 or 51, comprising a pump disposed between a reservoir for the emulsifying fluid and the emulsifying fluid channel.

[0447] Embodiment 53. A system as described in embodiment 52, including a controller configured to control the operation of the pump.

[0448] Embodiment 54. A system as described in embodiment 53, wherein the controller is configured to control the operation of the pump to achieve a target fluid velocity in the second microfluidic channel.

[0449] Embodiment 55. A system described in any of embodiments 32 to 54, including an imaging system positioned to capture an image of at least a portion of the chamber.

[0450] Embodiment 56. A system as described in embodiment 55, comprising a controller configured to control the flow rate of fluid through the inlet portion of the microfluidic channel based on an image captured by the imaging system.

[0451] Embodiment 57. The system described in embodiment 56, wherein the controller is configured to control the flow rate of the fluid by controlling the pressure applied to a reservoir fluidly connected to the inlet portion of the microfluidic channel.

[0452] Embodiment 58: A microfluidic chip that can be combined with any of embodiments 1 to 57, a plurality of first microfluidic channels for generating an emulsion of droplets of a first fluid in a second fluid, the first microfluidic channels being defined on a first surface of a microfluidic chip, each first microfluidic channel being fluidically independent from each other first microfluidic channel, each first microfluidic channel comprising: an inlet portion configured to receive the first fluid from a respective source of the first fluid; a junction between the inlet portion and a corresponding second fluid channel configured to carry a second fluid; and a plurality of first microfluidic channels including chambers downstream of the junctions, the chambers having cross-sectional areas greater than the cross-sectional areas of the inlet portions; a plurality of second microfluidic channels for polymerizing droplets of the emulsion, thereby producing a MOS, each second microfluidic channel fluidly connected to an outlet of a corresponding one of the first microfluidic channels; a plurality of second microfluidic channels, each of which is a serpentine channel including a first portion defined on a first surface of the microfluidic chip and a second portion defined on a second surface of the microfluidic chip opposite the first surface.

[0453] Embodiment 59. A microfluidic chip as described in embodiment 58, wherein each first microfluidic channel includes an outlet portion downstream of a chamber, and the cross-sectional area of ​​the chamber is larger than the cross-sectional area of ​​the outlet portion.

[0454] Embodiment 60. A microfluidic chip as described in embodiment 59, wherein a region of the outlet portion of each first microfluidic channel extends in a direction parallel to the respective chamber.

[0455] Embodiment 61. A microfluidic chip according to any one of embodiments 58 to 60, comprising a cover disposed on each of the first surface and the second surface of the microfluidic chip.

[0456] Embodiment 62. A microfluidic chip according to embodiment 61, wherein the cover comprises an optically transparent cover.

[0457] Embodiment 63. A microfluidic chip described in any of embodiments 58 to 62, wherein a plurality of first microfluidic channels are defined in a first region of the microfluidic chip and a plurality of second microfluidic channels are defined in a second region of the microfluidic chip different from the first region.

[0458] Embodiment 64. A microfluidic chip as described in embodiment 63, wherein one or more cutouts are defined between the first and second regions of the microfluidic chip.

[0459] Embodiment 65. A microfluidic chip as described in embodiment 64, wherein the edges of one or more cutouts are angled relative to the first and second surfaces of the microfluidic chip.

[0460] Embodiment 66. A microfluidic chip according to embodiment 64 or 65, wherein one or more cutouts extend through the entire thickness of the microfluidic chip.

[0461] Embodiment 67. A microfluidic chip according to any of embodiments 58 to 66, wherein each junction is a junction between a respective inlet portion and two corresponding second fluid channels.

[0462] Embodiment 68. A microfluidic chip according to embodiment 67, wherein the junction is a right-angle junction.

[0463] Embodiment 69. A microfluidic chip described in any of embodiments 58 to 68, wherein the microfluidic chip includes a plurality of inlet fingers, each inlet finger extending away from at least one other inlet finger and separated from each adjacent inlet finger by a gap, and at least some of the inlet portions of each first microfluidic channel being defined on the surface of the corresponding inlet finger.

[0464] Embodiment 70. A microfluidic chip described in any of embodiments 58 to 69, wherein the microfluidic chip includes a plurality of outlet fingers, each outlet finger extending away from at least one other outlet finger and separated from each adjacent outlet finger by a gap, and wherein the outlet portion of each second microfluidic channel is defined on the surface of the corresponding outlet finger.

[0465] Embodiment 71. A microfluidic chip described in any of embodiments 58 to 70, wherein the outlet port of each second microfluidic channel is configured to be connected to a corresponding cartridge for demulsifying an emulsion.

[0466] Embodiment 72. A device that can be combined with any of embodiments 1 to 71, 1. A cartridge for transferring MOS from an emulsion in a hydrophobic fluid to a suspension in an aqueous fluid, the demulsifying cartridge comprising: a collection reservoir defining a cavity for receiving a hydrophobic fluid; a substrate disposed on the collection reservoir, the substrate having a microfluidic channel defined in a first surface of the substrate facing the collection reservoir, the medium inlet channel for an aqueous fluid being fluidly connected to an upstream portion of the microfluidic channel; a hydrophobic membrane disposed between the collection reservoir and the surface of the substrate.

[0467] Embodiment 73. The device of embodiment 72, comprising a media reservoir having a cavity configured to contain an aqueous fluid, the media inlet channel being fluidly connected to the media reservoir.

[0468] Embodiment 74. The device of embodiment 73, comprising a tube extending through the substrate and the collection vessel, wherein the media inlet channel is fluidly connected to the media reservoir via the tube.

[0469] Embodiment 75. A device described in embodiment 73 or 74, wherein the collection vessel is disposed within the cavity of the media reservoir such that the collection vessel is positioned between the media reservoir and the substrate.

[0470] Embodiment 76. A device described in any of embodiments 73 to 75, wherein the bottom surface of the medium reservoir is angled relative to the plane of the substrate.

[0471] Embodiment 77. A device described in any of embodiments 73 to 76, comprising a duckbill valve disposed through an opening in the substrate and an opening in the collection container, the duckbill valve configured to allow aqueous fluid to be provided to the cavity of the medium reservoir.

[0472] Embodiment 78. A device described in any of embodiments 72 to 77, wherein the surface of the substrate is a first surface and a media inlet channel is defined on a second surface of the substrate opposite the first surface.

[0473] Embodiment 79. A device described in any of embodiments 72 to 78, wherein the cross-sectional area of ​​the microfluidic channel is larger at the upstream end of the microfluidic channel than at the downstream end of the microfluidic channel.

[0474] Embodiment 80. A device described in embodiment 79, wherein the upstream portion of the microfluidic channel has a simple serpentine shape and the downstream portion of the microfluidic channel has a double serpentine shape.

[0475] Embodiment 81. A device according to any of embodiments 72 to 80, comprising a hydrophobic absorbent material disposed within the cavity of the collection reservoir.

[0476] Embodiment 82. A method that can be combined with any of embodiments 1 to 81, generating droplets of a first fluid in a hydrophobic fluid in a droplet generation portion of a first microfluidic channel defined in a surface of the microfluidic chip, the first fluid comprising a biomaterial and a matrix material; applying a stimulus to the generated droplets in a polymerization portion of the first microfluidic channel to polymerize the matrix material, thereby forming an MOS emulsified in the hydrophobic fluid; transferring MOS from an emulsion to a suspension in an aqueous fluid, flowing a mixture of an aqueous fluid and an emulsion of MOS in a hydrophobic fluid along a second microfluidic channel defined in the substrate; and transferring the hydrophobic fluid through a membrane forming a wall of the second microfluidic channel as the mixture flows along the second microfluidic channel.

[0477] Embodiment 83. The method of embodiment 82, wherein generating droplets of the first fluid includes generating droplets at a junction between the first microfluidic channel and one or more channels carrying a hydrophobic fluid.

[0478] Embodiment 84. The method of embodiment 83, comprising controlling the flow rate of the hydrophobic fluid.

[0479] Embodiment 85. The method of any of embodiments 82-84, comprising controlling the flow rate of the first fluid based on the determined size of the generated droplets.

[0480] Embodiment 86. The method of embodiment 85, comprising determining the size of the generated droplets based on an image of the droplets in the droplet generation portion of the first microfluidic channel.

[0481] Embodiment 87. The method of any one of embodiments 82 to 86, wherein applying a stimulus to the generated droplets comprises heating the droplets.

[0482] Embodiment 88. The method of any of embodiments 82 to 87, wherein stimulating the generated droplets comprises irradiating the droplets with light having a wavelength configured to induce polymerization of the matrix material.

[0483] Embodiment 89. The method of embodiment 88, wherein the surface of the microfluidic chip is a first surface, the polymerized portion of the first microfluidic channel is defined on both the first surface and the second surface of the microfluidic chip, and irradiating the droplet includes irradiating the first surface and the second surface of the microfluidic chip.

[0484] Embodiment 90. The method of embodiment 88 or 89, wherein illuminating the droplets comprises illuminating the droplets with pulsed illumination.

[0485] Embodiment 91. The method of any of embodiments 82-90, comprising receiving the hydrophobic fluid to be transferred in a collection vessel, wherein a membrane is disposed between the collection vessel and the substrate.

[0486] Embodiment 92. The method of any of embodiments 82-91, wherein transferring the hydrophobic fluid through the membrane comprises applying a vacuum to the membrane.

[0487] Embodiment 93 The method of any of embodiments 82-92, comprising providing a suspension of MOS in an aqueous fluid to an effluent vial.

[0488] Embodiment 94. A method according to any of embodiments 82 to 93, comprising generating droplets of each of a plurality of first fluids in each of a plurality of fluidically independent first microfluidic channels defined on the surface of a microfluidic chip, and applying a stimulus to the generated droplets in each first microfluidic channel to form a MOS.

[0489] Embodiment 95. A method that can be combined with any of embodiments 1 to 94, comprising: flowing a first fluid through a first microfluidic channel of the microfluidic device, the first fluid comprising a biomaterial and a matrix material; flowing a second fluid through a second microfluidic channel of the microfluidic device, wherein the first fluid is immiscible with the second fluid; combining the first fluid and the second fluid to form droplets of the first fluid dispersed in the second fluid in a third channel of the microfluidic device; capturing, by an imaging device, a plurality of exposure images of the droplet of the first fluid in the third microfluidic channel within a single image captured by the imaging device, wherein capturing the plurality of exposure images includes: capturing, including illuminating, by a light source, a region of the third microfluidic channel with multiple successive illumination pulses during a single frame of the imaging device; determining a characteristic of the droplet based on an analysis of the captured exposure image; and controlling the flow of a first fluid in the first microfluidic channel, the flow of a second fluid in the second microfluidic channel, or both, based on the determined property of the droplet.

[0490] Embodiment 96. The method of embodiment 95, wherein flowing the second fluid into the second microfluidic channel comprises flowing the second fluid into two second microfluidic channels, and wherein confluence of the first fluid and the second fluid comprises confluence of the first fluid and the second fluid at a junction between the first microfluidic channel and the two second microfluidic channels.

[0491] Embodiment 97. The method of embodiment 96, comprising identifying droplets in each of the captured exposure images.

[0492] Embodiment 98. The method of embodiment 97, comprising identifying the leading edge of the droplet in each of the captured exposure images, the trailing edge of the droplet in each of the captured exposure images, or both.

[0493] Embodiment 99. The method of embodiment 97, comprising identifying droplets using frequency domain analysis or machine vision analysis, or by creating a best fit circle.

[0494] Embodiment 100. A method according to any of embodiments 96 to 99, wherein determining the characteristics of the droplet based on an analysis of the captured exposure image includes determining the distance traveled by the droplet between the time of a first one of the illumination pulses and the time of a second one of the illumination pulses.

[0495] Embodiment 101. A method according to any of embodiments 96 to 100, wherein determining the characteristics of the droplet based on analysis of the captured exposure image includes determining the velocity of the droplet in the third microfluidic channel.

[0496] Embodiment 102. A method according to any one of embodiments 96 to 101, wherein determining the characteristics of the droplets based on an analysis of the captured exposure image includes determining the size of the droplets.

[0497] Embodiment 103. A method according to any of embodiments 96 to 102, wherein determining the characteristics of the droplet based on analysis of the captured exposure image includes determining the distance between the droplet and an adjacent droplet in the third microfluidic channel.

[0498] Embodiment 104. A method according to any of embodiments 96 to 103, comprising determining an estimated number of droplets formed by the confluence of the first fluid and the second fluid based on the determined characteristics of the droplets.

[0499] Embodiment 105. The method of any of embodiments 96 to 104, comprising determining a droplet generation rate based on the determined characteristics of the droplets.

[0500] Embodiment 106. A method according to any of embodiments 96 to 105, comprising controlling the flow of a first fluid in a first microfluidic channel to obtain droplets of a target size.

[0501] Embodiment 107. A method according to any of embodiments 96 to 106, comprising controlling the flow of a second fluid in a second microfluidic channel to obtain droplets flowing at a target velocity through a third microfluidic channel.

[0502] Embodiment 108. The method of embodiment 107, wherein controlling the flow of the first fluid in the first microfluidic channel includes controlling the pressure of a fluid reservoir fluidically coupled to the first microfluidic channel, the fluid reservoir containing the first fluid.

[0503] Embodiment 109. The method of embodiment 107 or 108, wherein controlling the flow of the second fluid in the second microfluidic channel comprises operating a pump having a programmable flow rate to control the flow rate of the second fluid from a fluid reservoir fluidically coupled to the second microfluidic channel.

[0504] Embodiment 110. A method according to any one of embodiments 96 to 109, wherein acquiring multiple exposure images includes synchronizing the operation of a shutter of the imaging device with a light source of the imaging device.

[0505] Embodiment 111. The method described in embodiment 110, wherein synchronizing the operation of the shutter of the imaging device with the light source of the imaging device includes controlling the shutter of the imaging device to remain open while the region of the third microfluidic channel is illuminated with multiple consecutive illumination pulses.

[0506] Embodiment 112. A method as described in any of embodiments 96 to 111, wherein irradiating a region of the third microfluidic channel with multiple successive illumination pulses includes controlling the light source to emit illumination pulses each having a duration of 5 microseconds (μsec) to 125 μsec, for example, 25 to 50 μsec, and a pulse interval of about 1 millisecond (ms) to 50 ms, for example, 1 to 30 ms.

[0507] Embodiment 113. A method according to any of embodiments 96 to 112, wherein illuminating a region of the third microfluidic channel with multiple successive illumination pulses comprises illuminating a region of the third microfluidic channel with a first illumination pulse of a first color and a second illumination pulse of a second color.

[0508] Embodiment 114. A method according to any of embodiments 96 to 113, comprising illuminating a region of the third microfluidic channel with a collimated light source.

[0509] Embodiment 115. The method of any of embodiments 96-114, comprising exposing droplets of a first fluid to a stimulus sufficient to polymerize a matrix material in the droplets, thereby forming polymerized droplets dispersed in a second fluid.

[0510] Embodiment 116. The method of embodiment 115, comprising using polymerized droplets to assay a patient-specific therapy.

[0511] Embodiment 117: A system that can be combined with any of embodiments 1 to 116, a first microfluidic channel configured to be connected to a source of a first fluid; a second microfluidic channel configured to be connected to a source of a second fluid, wherein the first microfluidic channel and the second microfluidic channel intersect at a junction; a first controller configured to control flow regulators coupled to the first fluid source and the second fluid source; a third microfluidic channel downstream of the junction; and an imaging system including an imaging device and a light source; a second controller configured to control the imaging system to capture multiple exposure images of at least a portion of the third microfluidic channel within a single image captured by the imaging device; 1. A computing device including one or more processors coupled to a memory, the one or more processors coupled to the memory causing the computing device to: analyzing the plurality of exposure images of at least a portion of the third microfluidic channel to determine characteristics of the droplets in each of the plurality of captured exposure images; and a computing device configured to cause the first controller to control the flow regulator based on the determined characteristics of the droplet.

[0512] Embodiment 118. A system as described in embodiment 117, wherein the second controller is configured to control the shutter of the imaging device to open and to control the light source to generate multiple consecutive illumination pulses while the shutter of the imaging device is open.

[0513] Embodiment 119. A system described in embodiment 117 or 118, wherein one or more first controllers are configured to control a valve or pump for controlling the pressure of a first fluid reservoir fluidically coupled to a first microfluidic channel, the fluid reservoir containing a first fluid, and a pump having a programmable flow rate for controlling the flow rate of a second fluid from a second fluid reservoir fluidically coupled to a second microfluidic channel.

[0514] Embodiment 120. A system described in any of embodiments 117 to 119, wherein the light source includes a light-emitting diode.

[0515] Embodiment 121. A system described in any of embodiments 117 to 120, wherein the light source includes a collimated light source.

[0516] Embodiment 122. A system described in any of embodiments 117 to 121, wherein the light source includes multiple light sources, each configured to emit light of a different color.

[0517] Embodiment 123. A system described in any of embodiments 117 to 122, wherein one or more processors and memories are configured to cause the computing device to identify droplets in each of the captured exposure images.

[0518] Embodiment 124. A system as described in embodiment 123, wherein one or more processors and memories are configured to cause the computing device to identify the leading edge of the droplet in each captured exposure image, the trailing edge of the droplet in each captured exposure image, a best fit circle, or a combination thereof.

[0519] Embodiment 125. The system described in embodiment 124, wherein one or more processors and memories are configured to cause the computing device to identify droplets using frequency domain analysis or machine vision analysis.

[0520] Embodiment 126. A system described in any of embodiments 116 to 124, comprising a plurality of second microfluidic channels, wherein the first microfluidic channel and the plurality of second microfluidic channels intersect at a junction.

[0521] Embodiment 127. A system described in any of embodiments 117 to 126, wherein one or more processors and memories are configured to cause the computing device to determine the velocity of droplets in the third microfluidic channel.

[0522] Embodiment 128. A system described in any of embodiments 117 to 127, wherein one or more processors and memories are configured to cause the computing device to determine the size of droplets in the third microfluidic channel.

[0523] Embodiment 129. A system described in any of embodiments 117 to 128, wherein the system includes a fourth microfluidic channel connected to the third microfluidic channel, and a heating element positioned adjacent to the fourth microfluidic channel and configured to apply heat to at least a portion of the fourth microfluidic channel.

[0524] Embodiment 130: A system that can be combined with any of embodiments 1 to 129, a device configured to facilitate interaction between a first fluid stream and a second fluid stream within a flow path of the device; an optical sensor configured to acquire one or more images representative of the flow path; an image analysis module, processing the one or more images to identify at least one droplet generated in the flow path of the device by interaction between the first fluid stream and the second fluid stream; and an image analysis module configured to estimate a size of the at least one droplet; 1. A control system comprising: determining whether the size of at least one droplet satisfies a threshold condition; and a control system configured to generate a signal that makes an adjustment to a flow rate of at least one of the first fluid stream or the second fluid stream in response to a determination that the size of the at least one droplet satisfies a threshold condition.

[0525] Embodiment 131. The system described in embodiment 130, wherein the device is a microfluidic device.

[0526] Embodiment 132. A system described in embodiment 130 or 131, wherein the first fluid flow comprises a flow of a hydrophilic solution.

[0527] Embodiment 133. A system described in any of embodiments 130 to 130, wherein the first fluid flow comprises a flow of a solution comprising an unpolymerized mixture comprising a fluid matrix material and cells.

[0528] Embodiment 134. A system described in any of embodiments 130 to 133, wherein the second fluid flow comprises a flow of a hydrophobic solution.

[0529] Embodiment 135. A system described in any of embodiments 130 to 134, wherein the one or more images include an image of at least one droplet in a region of the device, and at least one droplet is not compressed by one or more walls of the device.

[0530] Embodiment 136. Processing one or more images acquired by an optical sensor includes: Detecting an edge of at least one droplet in at least one of the one or more images; identifying a first set of pixels corresponding to a detected edge of at least one droplet; identifying a circle corresponding to at least one droplet based on the first set of pixels; identifying a second set of pixels, the second set of pixels comprising a subset of the first set of pixels located within a threshold distance from the circumference of the identified circle; A system described in any of embodiments 130 to 135, comprising calculating a reference value representing the distance of at least a portion of the second set of pixels from a predetermined position within the at least one droplet.

[0531] Embodiment 137. The system described in embodiment 136, wherein processing one or more images acquired by the optical sensor further includes enhancing at least one of the one or more images after detecting the edge of at least one droplet.

[0532] Embodiment 138. A system described in embodiment 136 or 137, wherein processing one or more images acquired by the optical sensor further includes: shrinking at least one of the one or more images before identifying a first set of pixels corresponding to a detected edge of at least one droplet; and enlarging at least one of the one or more images after identifying a first set of pixels corresponding to a detected edge of at least one droplet.

[0533] Embodiment 139. A system described in any of embodiments 135 to 138, wherein the image analysis module is configured to estimate the size of at least one droplet based on a calculated reference value.

[0534] Embodiment 140. The system described in embodiment 139, wherein the calculated reference value includes a weighted average of the individual distances of a portion of the second set of pixels from a predetermined position within at least one droplet, and one or more weight values ​​of the weighted average are based on the intensity values ​​of the portion of the second set of pixels.

[0535] Embodiment 141. A system described in any of embodiments 136 to 140, wherein processing one or more images further includes excluding data corresponding to a detected edge of at least one droplet if one or more filtering conditions are met.

[0536] Embodiment 142. The system described in embodiment 141, wherein the one or more filtering conditions include an indication that the first set of pixels corresponds to a detected edge of a plurality of imaged droplets.

[0537] Embodiment 143. A system described in embodiment 141 or 142, wherein the one or more filtering conditions include an indication that the first set of pixels overlaps with at least one additional set of pixels.

[0538] Embodiment 144. A system described in any of embodiments 141 to 143, wherein the one or more filtering conditions include an indication of a detected signal that satisfies a threshold signal level condition, and the detected signal originates from within the perimeter of the first set of pixels.

[0539] Embodiment 145. A system described in any of embodiments 141 to 144, wherein the one or more filtering conditions include a determination that the first set of pixels is less than a proximity threshold from a wall of the imaged device.

[0540] Embodiment 146. A system described in any of embodiments 130 to 145, wherein the image analysis module is configured to estimate the size of at least one droplet with sub-pixel radial resolution.

[0541] Embodiment 147. A system described in any of embodiments 130 to 146, wherein the control system is configured to compare the size of at least one droplet with a target size obtained by user input.

[0542] Embodiment 148. A system described in any of embodiments 130 to 147, wherein the control system is configured to compare the size of at least one droplet with a target size automatically selected by the computing device.

[0543] Embodiment 149. A system described in any of embodiments 130 to 148, wherein the control system includes a feedback controller configured to generate a signal that adjusts the flow rate of at least one of the first fluid flow or the second fluid flow using proportional control, integral control, and / or derivative control.

[0544] Embodiment 150. A system described in any of embodiments 130 to 149, wherein the control system is configured to generate a signal to adjust the flow rate of at least one of the first fluid flow or the second fluid flow without feedback control if the image analysis module does not identify any droplets.

[0545] Embodiment 151. A system described in any of embodiments 130 to 150, wherein at least one droplet includes two or more droplets, and the control system is configured to increase or decrease the flow rate of at least one of the first fluid flow or the second fluid flow without feedback control when the standard deviation of the sizes of the two or more droplets exceeds a threshold value.

[0546] Embodiment 152. A system described in any of embodiments 130 to 151, wherein the image analysis module is further configured to process one or more images to identify air bubbles in the flow path.

[0547] Embodiment 153. A system described in any of embodiments 130 to 152, further comprising a storage device configured to store one or more images and / or data representing the size of at least one droplet.

[0548] Embodiment 154. A system described in any of embodiments 130 to 153, further comprising a transmission module configured to transmit one or more images and / or data representing the size of at least one droplet to a remote computing device.

[0549] Embodiment 155. A system described in any of embodiments 130 to 154, wherein the adjustment to the flow rate of at least one of the first fluid flow or the second fluid flow includes an adjustment to the pressure applied to the first fluid ...

Claims

1. 1. A microfluidic device comprising:

1. A microfluidic chip for generating a MicroOrganoSphere (MOS), comprising: a first microfluidic channel defined on a surface of the microfluidic chip; the first microfluidic channel comprising: a droplet generation portion comprising an inlet portion, a junction between the inlet portion and an emulsification fluid channel, and a chamber downstream of the junction, the chamber having a cross-sectional area greater than the cross-sectional area of ​​the inlet portion; a microfluidic chip comprising a polymerization section downstream of the droplet generation section, the polymerization section having a serpentine shape; A cartridge for MOS demulsification, comprising: collection containers, a substrate disposed on the collection reservoir, the substrate having a second microfluidic channel defined in a surface of the substrate facing the collection reservoir, the second microfluidic channel fluidly connected to an outlet of the polymerization portion of the first microfluidic channel; and a cartridge comprising a membrane disposed between the collection reservoir and the surface of the substrate.

2. 10. The microfluidic device of claim 1, wherein the droplet generation portion of the first microfluidic channel comprises an outlet portion downstream of the chamber, the cross-sectional area of ​​the chamber being greater than the cross-sectional area of ​​the outlet portion.

3. 10. The microfluidic device of claim 1, wherein the surface of the microfluidic chip is a first surface, and the polymerized portion of the microfluidic channel is defined on the first surface of the microfluidic chip and on a second surface of the microfluidic chip opposite the first surface.

4. 10. A microfluidic device according to any one of the preceding claims, wherein the junction comprises a junction with two hydrophobic fluidic channels.

5. 10. The microfluidic device of claim 1, wherein the membrane comprises a hydrophobic membrane.

6. 10. The microfluidic device of claim 1, wherein the surface of the substrate is a first surface, and a medium inlet channel is defined on a second surface of the substrate opposite the first surface of the substrate, the medium inlet channel being fluidly connected to an upstream section of the second microfluidic channel and configured to be connected to a medium reservoir.

7. 10. The microfluidic device of any one of the preceding claims, wherein the demulsifying cartridge comprises a hydrophobic material disposed within the collection reservoir.

8. 10. The microfluidic device of claim 1, wherein a vacuum channel is defined through a body of the collection reservoir, the vacuum channel configured to allow a vacuum to be applied to a surface of the membrane opposite the substrate.

9. 10. A microfluidic device according to any one of the preceding claims, comprising a reservoir fluidly connected to the first microfluidic channel via an inlet port defined at an inlet end of the first microfluidic channel.

10. 10. A microfluidic device according to any one of the preceding claims, wherein one or more cut-outs are defined between the droplet generation portion and the polymerization portion of the microfluidic chip.

11. 10. A microfluidic device according to any one of the preceding claims, wherein a plurality of first microfluidic channels are defined in the surface of the microfluidic chip, the device comprising a plurality of cartridges, the second microfluidic channel of each cartridge being fluidically connected to a corresponding one of the first microfluidic channels of the microfluidic chip.

12. 10. The microfluidic device of claim 1, wherein the device comprises an effluent vial fluidly connected to the second microfluidic channel via an outlet port defined in an outlet end of the second microfluidic channel.

13. 1. A system comprising: A microfluidic device according to any one of the preceding claims; a housing, the microfluidic device being disposed within the housing; and a polymerization block contained within the housing and positioned to apply a stimulus to the polymerization portion of the first microfluidic channel.

14. The system of claim 13 , wherein the polymerization block comprises a thermal polymerization block configured to apply heat to the polymerization portion of the first microfluidic channel.

15. 15. The system of claim 13 or 14, wherein the polymerization block comprises a photopolymerization block configured to illuminate the polymerized portion of the first microfluidic channel.

16. the surface of the microfluidic chip is a first surface, and the polymerized block is a first block disposed adjacent to the first surface of the microfluidic chip; a second block disposed adjacent to a second surface of the microfluidic chip, the second surface being opposite the first surface.

17. The system of any one of claims 13 to 16, comprising an imaging system positioned to capture an image of at least a portion of the chamber.

18. A microfluidic chip comprising: a plurality of first microfluidic channels for generating an emulsion of droplets of a first fluid in a second fluid, the first microfluidic channels being defined in a first surface of the microfluidic chip, each first microfluidic channel being fluidically independent from each other first microfluidic channel, each first microfluidic channel comprising: an inlet portion configured to receive the first fluid from a respective source of the first fluid; a junction between the inlet portion and a corresponding second fluid channel configured to carry the second fluid; and a plurality of first microfluidic channels comprising chambers downstream of the junctions, the chambers having cross-sectional areas greater than the cross-sectional areas of the inlet portions; a plurality of second microfluidic channels for polymerizing the droplets of the emulsion, thereby producing a MOS, each second microfluidic channel fluidly connected to an outlet of a corresponding one of the first microfluidic channels; a plurality of second microfluidic channels, each of which is a serpentine channel having a first portion defined on the first surface of the microfluidic chip and a second portion defined on a second surface of the microfluidic chip opposite the first surface.

19. 1. An apparatus comprising:

1. A cartridge for transferring MOS from an emulsion in a hydrophobic fluid to a suspension in an aqueous fluid, the cartridge comprising: a collection reservoir defining a cavity for receiving the hydrophobic fluid; a substrate disposed on the collection reservoir, the substrate having a microfluidic channel defined in a first surface of the substrate facing the collection reservoir, the substrate having a medium inlet channel for an aqueous fluid fluidically connected to an upstream portion of the microfluidic channel; a hydrophobic membrane disposed between the collection reservoir and the surface of the substrate.

20. 1. A method comprising: generating droplets of a first fluid in a hydrophobic fluid in a droplet generation portion of a first microfluidic channel defined in a surface of a microfluidic chip, the first fluid comprising a biomaterial and a matrix material; applying a stimulus to the generated droplets in a polymerization portion of the first microfluidic channel to polymerize the matrix material, thereby forming an emulsified MOS in the hydrophobic fluid; transferring the MOS from an emulsion to a suspension in an aqueous fluid, flowing a mixture of an aqueous fluid and the emulsion of MOS in the hydrophobic fluid along a second microfluidic channel defined in a substrate; and transferring the hydrophobic fluid through a membrane forming a wall of the second microfluidic channel as the mixture flows along the second microfluidic channel.

21. 1. A method comprising: flowing a first fluid through a first microfluidic channel of a microfluidic device, the first fluid comprising a biomaterial and a matrix material; flowing a second fluid through a second microfluidic channel of the microfluidic device, the first fluid being immiscible with the second fluid; combining the first fluid and the second fluid to form droplets of the first fluid dispersed in the second fluid in a third channel of a microfluidic device; capturing, with an imaging device, a plurality of exposure images of the droplet of the first fluid in the third microfluidic channel within a single image captured by the imaging device, wherein the capturing of the plurality of exposure images includes: capturing, including illuminating a region of the third microfluidic channel with multiple successive illumination pulses by a light source during a single frame of the imaging device; determining characteristics of the droplets based on an analysis of the captured exposure image; and controlling the flow of the first fluid in the first microfluidic channel, the flow of the second fluid in the second microfluidic channel, or both, based on the determined property of the droplet.

22. 22. The method of claim 21, comprising identifying the droplets in each of the captured exposure images.

23. 23. The method of claim 21 or 22, wherein determining characteristics of the droplet based on analysis of the captured exposure image comprises determining a distance traveled by the droplet between a time of a first one of the illumination pulses and a time of a second one of the illumination pulses.

24. The method of any one of claims 21 to 23, wherein determining a characteristic of the droplet based on an analysis of the captured exposure image comprises determining a size of the droplet.

25. 25. The method of any one of claims 21 to 24, comprising controlling the flow of the first fluid in the first microfluidic channel to obtain droplets of a target size.

26. 26. The method of claim 21, comprising controlling a second fluid in the second microfluidic channel, or both, to obtain droplets flowing through the third microfluidic channel at a target velocity.

27. 1. A system comprising: a first microfluidic channel configured to be connected to a source of a first fluid; a second microfluidic channel configured to be connected to a source of a second fluid, wherein the first microfluidic channel and the second microfluidic channel intersect at a junction; a first controller configured to control flow regulators coupled to the source of the first fluid and the source of the second fluid; a third microfluidic channel downstream of the junction; and an imaging system including an imaging device and a light source; a second controller configured to control the imaging system to capture multiple exposure images of at least a portion of the third microfluidic channel within a single image captured by the imaging device; and 1. A computing device comprising one or more processors coupled to a memory, the one or more processors coupled to the memory causing the computing device to: analyzing the plurality of exposure images of the at least a portion of the third microfluidic channel to determine droplet characteristics in each of the plurality of captured exposure images; and a computing device configured to cause the first controller to control the flow regulator based on the determined characteristic of the droplet.

28. 1. A system comprising: a device configured to facilitate interaction between a first fluid stream and a second fluid stream within a flow path of the device; an optical sensor configured to acquire one or more images representative of the flow path; an image analysis module, processing the one or more images to identify at least one droplet produced in a flow path of the device by the interaction between the first fluid stream and the second fluid stream; and an image analysis module configured to estimate a size of the at least one droplet; 1. A control system comprising: determining whether the size of the at least one droplet satisfies a threshold condition; and a control system configured to generate a signal to make an adjustment to a flow rate of at least one of the first fluid stream or the second fluid stream in response to a determination that the size of the at least one droplet satisfies the threshold condition.

29. processing the one or more images acquired by the optical sensor, detecting an edge of the at least one droplet in at least one of the one or more images; identifying a first set of pixels corresponding to the detected edge of the at least one droplet; identifying a circle corresponding to the at least one droplet based on the first set of pixels; identifying a second set of pixels, the second set of pixels comprising a subset of the first set of pixels located within a threshold distance from the circumference of the identified circle; and calculating a reference value representing a distance of at least a portion of the second set of pixels from a predetermined position within the at least one droplet.

30. 30. The system of claim 28 or 29, wherein the control system comprises a feedback controller configured to generate the signal that makes the adjustment to the flow rate of at least one of the first fluid flow or the second fluid flow using proportional control, integral control, and / or derivative control.

31. 1. A method comprising: one or more images representing a flow path within the microfluidic system that facilitates interaction between a first fluid stream and a second fluid stream; processing the one or more images to identify at least one droplet produced in the channel by the interaction between the first fluid stream and the second fluid stream; estimating a size of the at least one droplet; determining whether the size of the at least one droplet satisfies a threshold condition; generating a signal to adjust a flow rate of at least one of the first fluid stream or the second fluid stream in response to determining that the size of the at least one droplet satisfies the threshold condition.

32. 32. The method of claim 31 , wherein the signal is configured to increase or decrease the flow rate of at least one of the first fluid stream or the second fluid stream based on the size of the at least one droplet.

Images