Abstract: Organ segmentation in chest X-rays using convolutional neural networks is disclosed. One embodiment provides a method to train a convolutional segmentation network with chest X-ray images to generate pixel-level predictions of target classes. Another embodiment will also train a critic network with an input mask, wherein the input mask is one of a segmentation network mask and a ground truth annotation, and outputting a probability that the input mask is the ground truth annotation instead of the prediction by the segmentation network, and to provide the probability output by the critic network to the segmentation network to guide the segmentation network to generate masks more consistent with learned higher-order structures.

Abstract: A ceramic matrix composite (CMC) is formed by infiltrating a metal or alloy into a fiber preform in a reactor or furnace that is separated into multiple discrete temperature zones. The gradual cooling of the CMC is controlled, such that upon solidification, a narrow, planar, solidification front is created which allows the expanding metal or alloy to move into a hotter section of the fiber preform, opposed to the surface of the CMC. A discrete solidification front is established that moves through the ceramic matrix composite (CMC) as the composite cools.

Abstract: A method includes receiving topic names from a metrology interface device, where the topic names correspond to data generated by metrology devices associated with a manufacturing operation. The method also includes accessing a data stream from the metrology interface device. The data stream includes converted metrology data (in a common format) from the metrology devices and corresponding topic names. The method further includes extracting first converted metrology data from the data stream based on a first topic name of the topic names. The first converted metrology data is generated by converting first metrology data from a first metrology device of the metrology devices to the first converted metrology data having the common format. The method also includes comparing the first converted metrology data to a specification, detecting a condition based on the comparison, and sending a command to a manufacturing device based on the condition.

Abstract: Organ segmentation in chest X-rays using convolutional neural networks is disclosed. One embodiment provides a method to train a convolutional segmentation network with chest X-ray images to generate pixel-level predictions of target classes. Another embodiment will also train a critic network with an input mask, wherein the input mask is one of a segmentation network mask and a ground truth annotation, and outputting a probability that the input mask is the ground truth annotation instead of the prediction by the segmentation network, and to provide the probability output by the critic network to the segmentation network to guide the segmentation network to generate masks more consistent with learned higher-order structures.

Abstract: Organ segmentation in chest X-rays using convolutional neural networks is disclosed. One embodiment provides a method to train a convolutional segmentation network with chest X-ray images to generate pixel-level predictions of target classes. Another embodiment will also train a critic network with an input mask, wherein the input mask is one of a segmentation network mask and a ground truth annotation, and outputting a probability that the input mask is the ground truth annotation instead of the prediction by the segmentation network, and to provide the probability output by the critic network to the segmentation network to guide the segmentation network to generate masks more consistent with learned higher-order structures.

Abstract: A blade track assembly is disclosed. The blade track assembly comprises a plurality of ceramic matrix composite blade track segments arranged circumferentially adjacent to one another around a central axis. Each of the blade track segments includes an arcuate runner extending in a circumferential direction around a portion of the central axis and is coupled to circumferentially adjacent blade track segments.

Abstract: A method includes receiving topic names from a metrology interface device, where the topic names correspond to data generated by metrology devices associated with a manufacturing operation. The method also includes accessing a data stream from the metrology interface device. The data stream includes converted metrology data (in a common format) from the metrology devices and corresponding topic names. The method further includes extracting first converted metrology data from the data stream based on a first topic name of the topic names. The first converted metrology data is generated by converting first metrology data from a first metrology device of the metrology devices to the first converted metrology data having the common format. The method also includes comparing the first converted metrology data to a specification, detecting a condition based on the comparison, and sending a command to a manufacturing device based on the condition.

Abstract: A ceramic matrix composite (CMC) is formed by infiltrating a metal or alloy into a fiber preform in a reactor or furnace that is separated into multiple discrete temperature zones. The gradual cooling of the CMC is controlled, such that upon solidification, a narrow, planar, solidification front is created which allows the expanding metal or alloy to move into a hotter section of the fiber preform, opposed to the surface of the CMC. A discrete solidification front is established that moves through the ceramic matrix composite (CMC) as the composite cools.

Abstract: Organ segmentation in chest X-rays using convolutional neural networks is disclosed. One embodiment provides a method to train a convolutional segmentation network with chest X-ray images to generate pixel-level predictions of target classes. Another embodiment will also train a critic network with an input mask, wherein the input mask is one of a segmentation network mask and a ground truth annotation, and outputting a probability that the input mask is the ground truth annotation instead of the prediction by the segmentation network, and to provide the probability output by the critic network to the segmentation network to guide the segmentation network to generate masks more consistent with learned higher-order structures.

Abstract: A method of controlling the discharge of waste water from a toilet includes the installation of a valve in the discharge outlet of the toilet, the valve including a valve body and a spring-loaded valve closure member that is movable relative to the valve body into an open position during flushing of the toilet against the action of its spring-loading and which is returned to the closed position by the action of its spring-loading on completion of the flushing operation.

Abstract: A turbine shroud assembly or blade track assembly adapted to extend around a turbine wheel assembly is disclosed. The turbine shroud assembly includes a carrier and a blade track coupled to the carrier. The blade track is movable between a radially-inward position having a first inner diameter and a radially-outward position having a second inner diameter larger than the first inner diameter.

Abstract: A method for planning an infusion of a medical substance into a target region acquires non-patient-specific anatomical data from an atlas in the form of target region geometry information describing target and off-target regions for which an infusion of the medical substance is intended and regions to be avoided, respectively. Target coverage information is derived describing a minimum dose of the medical substance to be infused into the target region and describing a maximum dose of the medical substance to be applied to the off-target region. Relationship information describes a relationship between a physical parameter of an infusion setup and a selected spatial distribution of the medical substance. Infusion setup information for the infusion is determined that describes a planned infusion setup and is based on the relationship information and information related to the target and off-target regions.

Abstract: A blade track assembly is disclosed. The blade track assembly comprises a plurality of ceramic matrix composite blade track segments arranged circumferentially adjacent to one another around a central axis. Each of the blade track segments includes an arcuate runner extending in a circumferential direction around a portion of the central axis and is coupled to circumferentially adjacent blade track segments.

Abstract: A turbine shroud assembly or blade track assembly adapted to extend around a turbine wheel assembly is disclosed. The turbine shroud assembly includes a carrier and a blade track coupled to the carrier. The blade track is movable between a radially-inward position having a first inner diameter and a radially-outward position having a second inner diameter larger than the first inner diameter.

Abstract: The present invention refers to a data processing method for planning an infusion of a medical substance into a target region in an anatomical body part, wherein the medical substance is to be infused by means of an infusion setup which comprises an infusion device for performing the infusion as well as the medical substance, the method comprising: a) acquiring relationship data comprising relationship information which describes a relationship between at least one physical parameter of the infusion setup and a predetermined spatial distribution of the medical substance; b) acquiring target coverage data comprising target coverage information describing a planned coverage of the target region by the medical substance; c) determining infusion setup data comprising infusion setup information which describes a planned infusion setup, based on the relationship data and the target coverage data.

Abstract: An ink suitable for application to a heat resistant substrate and firing to fuse the ink to the substrate, the ink being in a form for ink jet printing and including a carrier material; a pigment; a fusible vitreous agent including particles of less than 10 microns in size, and the carrier having a melting point for phase change of the ink. The pigment includes ceramic pigment particles of less than 10 microns in size and the pigment and fusible vitreous agent are combined in the form of ceramic pigment particles. Dispersant is chemisorbed to these particles.

Abstract: An ink suitable for application to a heat resistant substrate and firing to fuse the ink to the substrate, the ink being in a form for ink jet printing and comprises; a carrier material; a pigment; a fusible vitreous agent comprising particles of less than 10 microns in size, and the carrier having a melting point for phase change of the ink. The pigment comprises ceramic pigment particles of less than 10 microns in size and the pigment and fusible vitreous agent are combined in the form of ceramic pigment particles. Dispersant is chemisorbed to these particles.