Abstract: The aorta and left atrium are localized from magnetic resonance data. The locations of the aorta and left atrium are detected jointly. The aorta and the left atrium are, at least in part, treated as one object. The detection may be from data representing a two-dimensional region. The two-dimensional region may be determined by first detecting the left ventricle from data representing a volume.

Abstract: An approach is provided for condition monitoring from log messages and sensor trends based on time semi-intervals. The approach may be applied to machine condition monitoring. Patterns are mined from symbolic interval data that extends previous approaches by allowing semi-intervals and partially ordered patterns. The semi-interval patterns and semi-interval partial order patterns are less restrictive than patterns using Allen's relations. Combinations and adaptations of efficient algorithms from sequential pattern and itemset mining for discovery of semi-interval patterns are described.

Abstract: A method for monitoring the quality-of-service (QoS) of high priority transactions in a software system includes receiving a specific QoS metric of a high priority transaction, where the QoS metric associated with a plurality of buckets and comparing the sampled specific QoS metric to an expected value for the specific QoS metric. If the sampled specific QoS metric exceeds the corresponding expected value, a bucket for the specific QoS metric is incremented, otherwise the bucket is decremented. If the bucket for the specific QoS metric overflows, the current bucket is reinitialized to zero, a depth of a next bucket for the specific QoS metric is computed, and a number of standard deviations from a mean value for the specific QoS metric is incremented. When the bucket for the specific QoS metric exceeds a threshold, a software rejuvenation routine is executed.

Abstract: A method and system for automatic extraction of personalized left atrium models is disclosed. A left atrium chamber body is segmented from a 3D image volume. At least one pulmonary venous ostium is detected on the segmented left atrium chamber body. At least one pulmonary vein trunk connected to the left atrium chamber body is segmented based on the detected pulmonary venous ostia.

Abstract: Adaptive demand response based on distributed load control may be provided. A demand response request indicative of a demand response event for an area may be received. A maximum power consumption for the area during the demand response event may be determined based on the demand response request. One or more set points for the area during the demand response event may be determined prior to the demand response event. The one or more set points are determined based on the maximum power consumption. Prior to or during the demand response event, a real-time power consumption for the area may be calculated. When the real-time power consumption exceeds the maximum power consumption, one or more loads in the area may be controlled to maintain the real-time power consumption at or below the maximum power consumption.

Abstract: A forensic video search user interface is disclosed that accesses databases of stored video event metadata from multiple camera streams and facilitates the workflow of search of complex global events that are composed of a number of simpler, low complexity events.

Abstract: Two adjacent objects with a gap between the objects rotate in a hot atmosphere with a temperature greater than 300 F in a gas turbine. Automatic and accurate contactless measurement of the gap is performed by taking images of the gap. An image, preferably an infra-red image is taken from the gap, a processor extracts the two edges from the image of the gap. The processor also determines a line through the pixels of an edge by applying a Hough transform on the pixels. The edges are substantially parallel. A line substantially perpendicular to the lines is also determined. Using the substantially parallel lines and the line substantially perpendicular to the substantially parallel lines the processor determines a width of the gap.

Abstract: Systems and methods for inspecting a device are disclosed. The method includes arranging the device in a known position relative to a plurality of movable cameras. The plurality of movable cameras is mounted on a controllable actuator. The plurality of cameras is pointed at the device by controlling the controllable actuator to position the camera with a user interface. An image of the device generated by the camera is displayed on a mobile and wireless display. The computing device also causes a rendered virtual image of the device to be displayed on the mobile and wireless display. A stream of interest and a region of interest is selected at the mobile and wireless display from the images generated by the cameras.

Abstract: A method for the detection of a balloon catheter within a fluoroscopic image, including: removing noise from a fluoroscopic image; detecting edges of a balloon catheter in the fluoroscopic image, wherein the detected edges include subsets of connected edges; extracting an edge subset from the subsets of connected edges; fitting a model to the extracted edge subset; removing outliers of the extracted edge subset based on the fitting of the model; adding the extracted edge subset without the outlier to a data set; repeating the extracting, fitting, removing and adding steps for the remainder of the subsets of connected edges; and fitting the model to the data set, wherein the data set is indicative of the balloon catheter.

Abstract: A method for device visualization includes receiving a set of physical characteristics including a description of spatial relationships of a plurality of markers within a device. Radiographic data of the device within a subject is acquired. An approximate location of each of the plurality of markers is identified within the radiographic data. A trajectory function is constructed for the device within the subject based on the identified approximate locations of each of the markers and the received set of physical characteristics. A section function is constructed for the device based on the set of physical characteristics and a 3D model is generated for the device based on the constructed trajectory function and the section function. A rendering of the 3D model is displayed on a display device.

Abstract: Localization of a coil is provided for magnetic resonance (MR)-guided intervention. A multi-scale decomposition and characteristic transitions in the power spectra for the coil are used to determine a distribution of likelihood of the coil being at each of various locations and/or to determine a confidence in the position determination. For example, the power spectra along each axis is used to generate a likelihood distribution of the location of the coil. The power spectra are decomposited at different scales. For each scale, the modulus maxima reflecting transitions in the power spectra are matched using various criteria. A likelihood is calculated for each of the matched candidates from characterizations of the matched candidates. The likelihood distribution is determined from a combination of the likelihoods from the various scales.

Abstract: A method and system for extracting rib centerlines in a 3D volume, such as a 3D computed tomography (CT) volume, is disclosed. Rib centerline voxels are detected in the 3D volume using a learning based detector. Rib centerlines or the whole rib cage are then extracted by matching a template of rib centerlines for the whole rib cage to the 3D volume based on the detected rib centerline voxels. Each of the extracted rib centerlines are then individually refined using an active contour model.

Abstract: A method for acquiring medical images, including: applying, during a first period, a plurality of radio frequency (RF) pulses to an area of interest, wherein the RF pulses applied during the first period are Kaiser-Bessel pulses; applying, during a second period, a plurality of 180 degree RF preparation pulses to the area; applying, during a third period, a plurality of 180 degree RF pulses to the area to acquire a center of a k-space; applying, during a fourth period, a plurality of RF pulses to the area, wherein the RF pulses applied during the fourth period have an angle smaller than the 180 degree RF pulses applied during the third period; applying, during a fifth period, a plurality of constant RF pulses to the area to acquire outer lines of the k-space; and generating an image of the area by using a steady-state free precession echo readout.

Abstract: A method (10, 100) for temporal encoding for digital subtraction angiography (DSA) and other angiographic data.

Abstract: A method for training a classifier for selecting features in sparse data sets with high feature dimensionality includes providing a set of data items x and labels y, minimizing a functional of the data items x and associated labels y L ? ( w , b , a , c , ? 1 , ? 2 ) := 1 N ? ? i = 1 N ? a i + ? 1 ? ? c ? 1 + ? 2 2 ? ? w ? 2 2 + ? 1 T ? ( e - Y ? ( Xw + be ) - a ) + ? 2 T ? ( w - c ) + ? 1 2 ? ? e - Y ? ( Xw + be ) - a ? 2 2 + ? 2 2 ? ? w - c ? 2 2 to solve for hyperplane w and offset b of a classifier by successively iteratively approximating w and b, auxiliary variables a and c, and multiplier vectors ?1 and ?2, wherein ?1, ?2, ?1, and ?2 are predetermined constants, e is a unit vector, and X and Y are respective matrix representations of the data items x and labels y; providing non-zero elements of the hyperplane vector w and corresponding components of X and Y as arguments to an interior point method solv

Abstract: A method and system for up-vector detection for ribs in a 3D medical image volume, such as a computed tomography (CT) volume is disclosed. A rib centerline of at least one rib is extracted in a 3D medical image volume. An up-vector is automatically detected at each of a plurality of centerline points of the rib centerline of the at least one rib. The up-vector at each centerline point can be detected using a trained regression function. Alternatively, the up-vector at each centerline point can be detected by detecting an ellipse shape in a cross-sectional rib image generated at each centerline point.

Abstract: A method and system for automatic magnetic resonance (MR) volume composition and normalization is disclosed. In one embodiment, a plurality of MR volumes is received. A composite MR volume is generated from the plurality of MR volumes. Volume normalization of the composite MR volume is then performed to correct intensity inhomogeneity in the composite MR volume. The volume normalization of the composite MR volume may be performed using template MR volume or without a template MR volume.

Abstract: A method (100) that registers a 3D heart volume (112, 114) obtained from either a pre-operative MR image or CT image (102) to an intra-operative fluoroscopic image using a mesh of the heart structure (106) as the basis for the registration.

Abstract: An object in a hot atmosphere with a temperature greater than 400 F in a gas turbine moves in a 3D space. The movement may include a vibrational movement. The movement includes a rotational movement about an axis and a translational movement along the axis. Images of the object are recorded with a camera, which may be a high-speed camera. The object s provided with a pattern that is tracked in images. Warpings of sub-patches in a reference image of the object are determined to form standard format warped areas. The warpings are applied piece-wise to areas in following images to create corrected images. Standard tracking such as SSD tracking is applied to the piece-wise corrected images to determine a movement of the object. The image correction and object tracking are performed by a processor.

Abstract: Automatic prostate localization in T2-weighted MR images facilitate labor-intensive cancer imaging techniques. Methods and systems to accurately segment the prostate gland in MR images are provided and address large variations in prostate anatomy and disease, intensity inhomogeneities, and artifacts induced by endorectal coils. A center of the prostate is automatically detected with a boosted classifier trained on intensity based multi-level Gaussian Mixture Model Expectation Maximization (GMM-EM) segmentations of the raw MR images. A shape model is used in conjunction with Multi-Label Random Walker (MLRW) to constrain the seeding process within MLRW.