Siemens Medical Solutions Usa, Inc.

Siemens Healthineers Ag

Bimba Rao

Ismayil M Guracar

Helene C Houle

Bogdan Georgescu

Tommaso Mansi

Kutay F Ustuner

Ingmar Voigt

Rui Liao

Viorel Mihalef

Danhua Zhao

Seshadri Srinivasan

Bruce A McDermott

Liang Zhai

Yue Zhang

Abdoul Aziz Amadou

Henry G Pavy

Sharleen Marshall

Miroslav Gajdos

Cynthia Wolverton

Matthias John

Stephen Henderson

Anand Tatpati

For a classification-dependent user interface in ultrasound imaging with an ultrasound scanner, the ultrasound scanner classifies a view represented in an image. The user interface changes according to the view, allowing one or a few user inputs to be used for different user options or behavior combinations appropriate for the classified anatomy. The context from imaging alters the behavior of a given user input element of the user interface.

Image classification-dependent user interface in ultrasound imaging

A method, apparatus, and computer readable storage medium are provided herein for constructing a representation of an annular structure associated with an anatomical object. The method includes receiving three-dimensional image data of the anatomical object and detecting at least a first landmark point and a second landmark point on the annular structure. A plane positioned between the first landmark point and the second landmark point, and oriented in accordance with a predefined angular relationship to a line connecting the first landmark point and the second landmark point is determined. A third landmark point on the annular structure which lies in the plane is also detected and the representation of the annular structure is generated using at least the first landmark point, the second landmark point, and the third landmark point. The representation is then outputted.

Annular structure representation

Machine learning trains to tune settings. For training, the user interactions with the image parameters (i.e., settings) as part of ongoing examination of patients are used to establish the ground truth positive and negative examples of settings instead of relying on an expert review of collected samples. Patient information, location information, and/or user information may also be included in the training data so that the network is trained to provide settings for different situations based on the included information. During application, the patient is imaged. The initial or subsequent image is input with other information (e.g., patient, user, and/or location information) to the machine-trained network to output settings to be used for improved imaging in the situation.

Tuned medical ultrasound imaging

Using computer-assisted classification and/or computer-assisted segmentation with or without monitoring the field of view for change, the workflow for ultrasound imaging may be made more efficient. The classification and/or segmentation is used to perform a next act in the sequence of acts making up the ultrasound examination. Rather than requiring a user to determine the act and implement the act, the ultrasound scanner determines and implements the act based on the identification and/or location of an imaged object. For example, the identification of the object as a kidney using a machine-learnt classifier triggers a color flow scan, and the location of the object determines a placement for the color flow region of interest (ROI), avoiding the user having to perform the ROI initiation and/or placement and increasing workflow efficiency.

Machine-aided workflow in ultrasound imaging

Image enhancement is provided for medical diagnostic ultrasound. Knowledge-based detection of anatomy or artifact identifies locations to be enhanced. The knowledge-based detection of the locations may avoid identification of other anatomy or artifacts. The image enhancement is applied to the identified locations and not others.

Knowledge-based ultrasound image enhancement

Transmit power is adaptively set in medical diagnostic ultrasound imaging. The relative strength, such as a ratio, of harmonic and fundamental responses is calculated. This relative strength is used to set the transmit power. The transmit power may be set following ALARA while providing enough information for harmonic imaging.

Transmit power based on harmonic to fundamental relationship in medical ultrasound imaging

An entire volume is scanned. A sub-volume is separately scanned with different settings for beamforming parameters, allowing greater image quality for the sub-volume while providing context from the volume. The anatomy of interest is periodically detected, and the sub-volume shifted in position to cover the anatomy of interest, allowing for relatively continuous volume imaging with enhanced quality imaging of the sub-volume. Interleaving by volume and sub-volume slices may allow for optimization of relative frame rate and image quality. Different combinations between volume and sub-volume data for anatomy detection and display may provide for desired imaging while allowing the regular detection of the anatomy.

Continuously oriented enhanced ultrasound imaging of a sub-volume

For volume ultrasound imaging, the view direction for rendering a volume image is used. In one approach, imaging parameters for acquiring the volume data prior to rendering are set based on the view direction. For the resolution example, the data for imaging is acquired with the resolution in a view plane greater than resolution along the ray tracing direction. In another approach, sets of data representing a volume are acquired or formed with different settings. The sets are weighted based on the view direction and combined. For the resolution example, an image rendered from the combined data may have greater resolution in the view plane where the set of data with greater resolution in that view plane is weighted more in the combination.

View direction adaptive volume ultrasound imaging

Steered spatial compounding is provided in ultrasound imaging. Component frames of data associated with different spatial response are processed to identify different sources of signals. Matched filtering and/or spatial variance with erosion distinguish between sources of signals, such as distinguishing between (1) anatomical structure and (2) soft tissue and/or noise. The locations in the patient are segmented by the source of signals and different compounding is applied for the different sources.

Component frame enhancement for spatial compounding in ultrasound imaging

Image enhancement is provided in ultrasound imaging. A filter bank is used to process the data. The data may be processed in parallel. Each filter outputs data representing a same location, but with a kernel path (e.g., line, curved line, area, curved surface, or volume) different than for the other filters. For example, three filters filter along respective orthogonal planes. The filters may use preselected or data independent paths. The output of the filter bank is determined from the outputs of the filters in the bank, such as by selecting the maximum value. This bank-based filtering may enhance data associated with anatomical surfaces or contours and/or may reduce speckle. The parallel operation of the filters may allow for more rapid processing and/or processing a greater amount of data as compared to using a single filter.

Filter bank for ultrasound image enhancement

Orientation is automatically identified in medical diagnostic ultrasound image. An area or volume based process, such as region shrinking or using locations associated with flow or tissue structure, determines a direction or orientation. Real-time imaging in B-mode and/or flow mode may be enhanced based on the direction or orientation.

Automatic identification of orientation in medical diagnostic ultrasound

Image processing adapts to speckle. Speckle is identified from signal transitions. For example, peaks, valleys or mean crossings of image signals as a function of space or spatial location are identified. A speckle characteristic, such as speckle size, is estimated from the signal transitions. The estimation may be limited to soft tissue regions to reduce the effects of specular targets and noise on speckle estimation. The speckle is estimated for local regions or an entire image. By estimating speckle for local regions, image processing may account adaptively for regional variation in speckle size.

Speckle adaptive medical image processing

Clinical markers are adaptively identified in steered spatial compounding ultrasound imaging. Motion is detected. The use or determination of the clinical marker information adapts to the amount of motion. If sufficient motion is detected, the contribution of clinical marker information may be reduced, the weighting of the clinical marker information relative to the steered compound information is reduced, the acquisition sequence is adjusted for component frames of data, or combinations thereof.

Growing community of inventors

San Jose, CA, United States of America

Bimba Rao