Siemens Medical Systems, Inc.

Maplebear Inc.

Ming Li

Jin H Kim

Dong-Chyuan Liu

Levin F Nock

Zhiyi Fan

Dmytro Solomadin

Natalie Binns

An online system receives information identifying items associated with a brand, a hierarchical taxonomy of the items, and information identifying a retailer associated with the brand. The system applies a machine learning model to predict availabilities of the items at (a) retailer location(s) associated with the retailer, identifies items that are likely available at the retailer location(s), and groups the identified items into categories based on the taxonomy. The system computes an item score for each item based on its popularity, attributes, and/or attributes of a user. The system assigns items in each category to positions within a display unit associated with the category and computes a category score for each category based on the item scores. The system assigns display units associated with the categories to positions within a template based on the category score and generates a page associated with the brand and retailer based on the assignments.

Automatically generating a retailer-specific brand page based on a machine learning prediction of item availability

An ultrasonic imaging system has a three-dimensional acoustic display. Beam-formed ultrasound scan signals are separated into channel signals and are then filtered using head-related transfer function (HRTF) filters. An HRTF is associated with each of several points in an interrogation region. A filter coefficient processor selects, for each channel signal, a set of HRTF filter parameters according to which sample volume the channel signal comes from. Each HRTF parameter set defines left-ear and right-ear response functions to a calibration sound generated at a calibration point in a calibration space; the calibration point itself corresponds to a predetermined position in the interrogation region. The HRTF-filtered channel signals are then combined into left and right audio output signals which drive left and right speakers that the user listens to while performing the ultrasound scan. The calibration points in the calibration space are mapped to respective reference sample volumes located in a predetermined ultrasonic interrogation region.

Acoustic display system and method for ultrasonic imaging

A region of tissue is scanned as a series of 2-D frames. The correlation of substantially homogeneous speckle regions in the frames is determined and is compared with a precalibrated speckle correlation-versus-distance function to give an estimate of the distance between the frames. In some applications, the frames are partitioned into sub-frames, whose correlation values and distances are combined according to predetermined functions to give a distance estimate between the frames as a whole. The reliability of the distance estimates is improved and evaluated in various ways, for example, by comparing combinations of possible distances from end frames via intermediate frames, and by comparing computed frame or sub-frame velocities with a known or estimated transducer velocity. Once the relative distances of the 2-D frames are estimated, a 3-D image is compiled using registration techniques. Image frames need not be parallel.

System and method for 3-D medical imaging using 2-D scan data

An ultrasonic imaging system is provided that effectively corrects for inhomogeneities in the imaged media. The ultrasonic imaging system comprises a transducer array having a plurality of transducer elements each being separately operable to provide an ultrasonic pulse in response to a driving signal and to provide a corresponding echo signal in response to a received echo based on the ultrasonic pulse. A transmitter is coupled to the transducer array and provides the driving signal to each respective one of the transducer elements. The driving signals define a steered beam of the transducer array. A receiver is coupled to the transducer array and receives the echo signal from each respective one of the transducer elements. The receiver includes a beamformer section that produces a single matched and summed signal from the echo signals of the plurality of transducer elements. The beamformer section comprises a plurality of processing modules coupled in a cascading arrangement with separate processing stages.

Beamformer for phase aberration correction

An interrogation region of a patient's body is scanned with an ultrasonic probe. At least one input data frame is generated representing the scanned interrogation region as a pattern of image elements, each with a corresponding input intensity value, that are stored in a memory. The input data frame is then displayed as a pattern of displayed intensity values, which the processor calculates as combination of a geometry-independent function of an ambient light input array and a geometry-dependent function of the input intensity values, in which the geometry-dependent function includes a coefficient term that multiplies the input intensity array and that itself is a function of the input intensity values. As part of the coefficient term is a filtered input data array, which results from noise-floor filtering, low-pass filtering, and surface normal vector averaging of the input intensity array. In a preferred embodiment of the invention, the user selects an assumed light source vector, and/or an assumed viewer direction vector, which are the incorporated into the geometry-dependent function used to calculate displayed intensity values.

System and method for topographic mapping of ultrasonic image displays

A method for measuring the interframe movement of pixels in a second ultrasoud image frame with respect to the pixels in a first ultrasoud image frame. The movement information may be displayed to provide enhanced information about the underlying physiology or to provide a temporally compounded image having reduced noise. In the preferred embodiment of the present invention, the pixels in that actually change between frames are identified. The first frame is divided into blocks and an estimated velocity computed for each block having a sufficient number of pixels that actually moved between the first and second frames. After all of the velocity estimates have been computed, blocks having questionable velocity estimates are identified. Blocks for which the velocity estimate was questionable are then provided with a velocity estimate by interpolating the velocity vectors of the nearby blocks having valid velocity estimates. The preferred method for estimating the velocity utilizes a modified optical flow computation that is more stable than prior art optical flow models.

Noise suppression method utilizing motion compensation for ultrasound

Apparatus for adaptive phase aberration correction of wave distortions caused by tissue inhomogeneities in ultrasound images. In accordance with the present invention, auto-correlation apparatus in multiple, parallel, color flow processors compute a cross-correlation which is used to correct distortions in real time.

Growing community of inventors

Seattle, WA, United States of America

Ming Li