Abstract: A microfabricated droplet dispensing structure is described, which may include a MEMS microfluidic fluidic valve, configured to open and close a microfluidic channel. The opening and closing of the valve may separate a target biological particle containing genomic material, and a bead from a sample stream, and direct these two particle into a single droplet formed at the edge of the substrate. The droplet may then be encased in a sheath flow of an immiscible fluid, and provided to a sequencing module. The sequencing module may sequence the genomic material and/or an identifying barcode attached to the bead.

Abstract: A particle manipulation system uses a MEMS-based, microfabricated particle manipulation device which has a sample inlet channel, output channels, and a movable member formed on a substrate. The device may be used to separate a target particle from non-target material in a sample stream. In order to improve the sorter speed, accuracy or yield, the particle manipulation system may also include a microfluidic structure which focuses the target particles in a particular portion of the sample inlet channel. The multiple outputs may be implemented using structures outside of, and in addition to, the microfabricated sorter. These may include pneumatic valves, which may transfer the contents of one fluid receptacle to another. This may enable simple and inexpensive serial sorting.

Abstract: A cell sorting device is disclosed, wherein the device includes a sorting magnet and at least one particle manipulation device, wherein the particle manipulation device is formed on a surface of a fabrication substrate. The device may include at least one fluid channel, wherein the sorting magnet and the particle manipulation device are in fluid communication with one another through at least one fluid channel. A method of sorting cells from a first cell suspension is also disclosed, The method may include a) magnetic labeling of first target cells and removal of the non-target cells by applying magnetic fields to obtain a second cell suspension; b) fluorescence-activated labeling of second target cells present in the second cell suspension and separating the fluorescence-activated second target cells from the not labeled cells to obtain a third cell suspension.

Abstract: A MEMS-based particle manipulation system which uses a particle manipulation stage and a plurality of laser interrogation regions. The laser interrogation regions may be used to assess the effectiveness or accuracy of the particle manipulation stage. In one exemplary embodiment, the particle manipulation stage is a microfabricated, flap-type fluid valve, which sorts a target particle from non-target particles in a fluid stream. The laser interrogation stages are disposed in the microfabricated fluid channels at the input and output of the flap-type sorting valve. The laser interrogation regions may be used to assess the effectiveness or accuracy of the sorting, and to control or adjust sort parameters during the sorting process. One or more feedback loops may be used to improve the particle manipulation process, based on data acquired during the first interrogation and/or during a downstream confirmation. Artificial intelligence techniques may be used to good effect.

Abstract: A microfabricated droplet dispensing structure is described, which may include a MEMS microfluidic fluidic valve, configured to open and close a microfluidic channel. The opening and closing of the valve may separate a target biological particle containing genomic material, and a bead from a sample stream, and direct these two particle into a single droplet formed at the edge of the substrate. The droplet may then be encased in a sheath flow of an immiscible fluid, and provided to a downstream workflow.

Abstract: A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Abstract: A microfabricated particle manipulation system, wherein a target particle is pierced by a microfabricated actuator or by a microfabricated knife edge. In either case, the particle membrane is altered, so as to allow material to traverse the membrane. The device may be used to extract cellular material from inside a cell, or to transfect a cell with foreign material.

Abstract: A microfabricated droplet dispensing structure is described, which may include a MEMS microfluidic fluidic valve, configured to open and close a microfluidic channel. The opening and closing of the valve may separate a target biological particle containing genomic material, and a bead from a sample stream, and direct these two particle into a single droplet formed at the edge of the substrate. The droplet may then be encased in a sheath flow of an immiscible fluid, and provided to a sequencing module. The sequencing module may sequence the genomic material and/or an identifying barcode attached to the bead.

Abstract: A microfabricated droplet dispensing structure is described, which may include a MEMS microfluidic fluidic valve, configured to open and close a microfluidic channel. The opening and closing of the valve may separate a target particle and a bead from a sample stream, and direct these two particle into a single droplet formed at the edge of the substrate. The droplet may then be encased in a sheath flow of an immiscible fluid.

Abstract: A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Abstract: A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Abstract: A particle manipulation system uses a MEMS-based, microfabricated particle manipulation device which has a sample inlet channel, output channels, and a movable member formed on a substrate. The device may be used to separate a target particle from non-target material in a sample stream. In order to improve the sorter speed, accuracy or yield, the particle manipulation system may also include a microfluidic structure which focuses the target particles in a particular portion of the sample inlet channel. The particle manipulation device may have two separate sort output channels, wherein the sort channel used depends on the characteristics of the sort pulse delivered to the micromechanical particle manipulation device.

Abstract: A cell sorting device is disclosed, wherein the device includes a sorting magnet and at least one particle manipulation device, wherein the particle manipulation device is formed on a surface of a fabrication substrate. The device may include at least one fluid channel, wherein the sorting magnet and the particle manipulation device are in fluid communication with one another through at least one fluid channel. A method of sorting cells from a first cell suspension is also disclosed, The method may include a) magnetic labeling of first target cells and removal of the non-target cells by applying magnetic fields to obtain a second cell suspension; b) fluorescence-activated labeling of second target cells present in the second cell suspension and separating the fluorescence-activated second target cells from the not labeled cells to obtain a third cell suspension.

Abstract: A microfabricated droplet dispensing structure is described, which may include a MEMS microfluidic fluidic valve, configured to open and close a microfluidic channel. The opening and closing of the valve may separate a target particle and a bead from a sample stream, and direct these two particle into a single droplet formed at the edge of the substrate. The droplet may then be encased in a sheath flow of an immiscible fluid.

Abstract: A microfabricated particle manipulation system, wherein a target particle is pierced by a microfabricated actuator or by a microfabricated knife edge. In either case, the particle membrane is altered, so as to allow material to traverse the membrane. The device may be used to extract cellular material from inside a cell, or to transfect a cell with foreign material.

Abstract: A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Abstract: Described here is a microfabricated particle sorting device that uses a transient pulse of fluidic pressure to deflect the target particle. The transient pulse may be generated by a microfabricated (MEMS) actuator, which pushes a volume of fluid into a channel, or sucks a volume of fluid from the channel. The transient pressure pulse may divert a target particle into a sort channel.

Abstract: Described here is a microfabricated particle filtering structure having at least one microchannel formed in a surface of a silicon substrate. The filter structure uses a plurality of barriers formed in at least one microchannel, wherein a distance between the closest barriers is small enough to capture the particulate debris but allow the sample fluid to flow, and a transparent layer that covers the silicon substrate, and the plurality of barriers. The debris captured by the barriers may be analyzable through the transparent layer, helping in determining the source of the debris.

Abstract: Described here is a microfabricated particle sorting device that uses a transient pulse of fluidic pressure to deflect the target particle. The transient pulse may be generated by a microfabricated (MEMS) actuator, which pushes a volume of fluid into a channel, or sucks a volume of fluid from the channel. The transient pressure pulse may divert a target particle into a sort channel.

Abstract: A particle manipulation system uses a MEMS-based, microfabricated particle manipulation device which has a sample inlet channel, output channels, and a movable member formed on a substrate. The device may be used to separate a target particle from non-target material in a sample stream. In order to improve the sorter speed, accuracy or yield, the particle manipulation system may also include a microfluidic structure which focuses the target particles in a particular portion of the sample inlet channel. This focusing element may include cavities of variable cross section along the channel length. In addition, a filtering element may also be included upstream of the focusing element.