Abstract: Aspects of the technology described herein relate to methods and apparatuses for enable a user to manually modify an input to a calculation performed based at least in part on an ultrasound image. In some embodiments, manually modifying the input comprises manually modifying a trace on the image. In some embodiments, manually modifying the input comprises manually selecting an ultrasound image from a series of ultrasound images on which a calculation is to be performed.

Abstract: Aspects of the technology described herein relate to techniques for receiving feedback from a user regarding an automatic calculation performed based on ultrasound data. The automatic calculation may be a result for a measurement performed on the ultrasound data. The feedback from the user may include an indication of agreement or disagreement with the result of the measurement; an indication of whether the result of the measurement is too high, too low, or correct; a result for the measurement that the user considers to be correct; or locations on one or more ultrasound images where one or more statistical models should have focused when performing the automatic calculation. The automatic calculation may also include a quality of the ultrasound data for performing a measurement, and the feedback may include an indication whether the user considers the ultrasound data acceptable for performing the measurement or not.

Abstract: Aspects of the technology described herein relate to a processing device in operative communication with an ultrasound device, where the processing device is configured to capture video with a front-facing camera and display, simultaneously, the video and an instruction for moving the ultrasound device. The video may depict the ultrasound device and portions of the user near the ultrasound device. The processing device may further display, simultaneously with the video and the instruction, an ultrasound image generated based on ultrasound data received from the ultrasound device. The instruction may be an instruction for moving the ultrasound device from a current position and orientation relative to the user to a target position and orientation relative to the user.

Abstract: Aspects of the technology described herein relate to determining, by a processing device in operative communication with an ultrasound device, a position and/or orientation of the ultrasound device relative to the processing device, and displaying on a display screen of the processing device, based on the position and/or orientation of the ultrasound device relative to the processing device, an instruction for moving the ultrasound device.

Abstract: A method of determining the position and orientation of an invasive medical device is described using altered devices and a machine learning algorithm for detecting the position and orientation of such devices from imagery. Such predictions can subsequently be displayed to an operator to improve the speed and accuracy by which they perform procedures.

Abstract: Aspects of the technology described herein relate to displaying indications of anatomical regions that have been imaged and/or that should be imaged next. In some embodiments, ultrasound data collected from a subject by an ultrasound device may be received, an automatic determination may be made that the ultrasound data was collected from a particular anatomical location, and an indication of the anatomical location may be displayed. In some embodiments, a plurality of anatomical locations may be determined for imaging, an anatomical location from among the plurality of anatomical locations may be automatically selected, and an indication of the anatomical location may be displayed. Displaying an indication of an anatomical location may include displaying or modifying a marker on a frame of a video of the subject such that the marker appears in the frame of the video to be located at the anatomical location on the subject.

Abstract: Aspects of the technology described herein relate to displaying locations on images of body portions. Based on ultrasound data collected from a subject by an ultrasound device, a first location on an image of a body portion may be determined. The first location on the image of the body portion may correspond to a current location of the ultrasound device relative to a subject where the ultrasound device collected the ultrasound data. A first marker may be displayed on the image of the body portion at the first location. A second location on the image of the body portion may be determined, where the second location on the image of the body portion corresponds to a target location of the ultrasound device relative to the body portion of the subject. A second marker on the image of the body portion at the second location may be displayed.

Abstract: Aspects of the technology described herein relate to guiding collection of ultrasound data collection using motion and/or orientation data. A first instruction for rotating or tilting the ultrasound imaging device to a default orientation may be provided. Based on determining that the ultrasound imaging device is in the default orientation, a second instruction for translating the ultrasound imaging device to a target position may be provided. Based on determining that the ultrasound imaging device is in the target position, a third instruction for rotating or tilting the ultrasound imaging device to a target orientation may be provided.

Abstract: Aspects of the technology described herein relate to guiding collection of ultrasound data collection using motion and/or orientation data. Motion and/or orientation data may be received from an ultrasound imaging device, where the motion and/or orientation data provides an indication of the motion and/or orientation of the ultrasound imaging device. Ultrasound data collected by the ultrasound imaging device may also be received. An instruction for moving the ultrasound imaging device based on the motion and/or orientation data and the ultrasound data may be provided. Ultrasound data and motion and/or orientation data may indicate a velocity of the ultrasound imaging device that exceeds a threshold velocity and an instruction for slowing the velocity of the ultrasound imaging device may be provided.

Abstract: A method of determining the position and orientation of an invasive medical device is described using altered devices and a machine learning algorithm for detecting the position and orientation of such devices from imagery. Such predictions can subsequently be displayed to an operator to improve the speed and accuracy by which they perform procedures.

Abstract: Aspects of the technology described herein relate to configuring an ultrasound system with imaging parameter values. In particular, certain aspects relate to configuring an ultrasound system to produce a plurality of sets of ultrasound images, each respective set of the plurality of sets of ultrasound images being produced with a different respective set of a plurality of sets of imaging parameter values; obtaining, from the ultrasound system, the plurality of sets of ultrasound images; determining a set of ultrasound images from among the plurality of sets of ultrasound images that has a highest quality; and based on determining the set of ultrasound images from among the plurality of sets of ultrasound images that has the highest quality, automatically configuring the ultrasound system to produce ultrasound images using a set of imaging parameter values with which the set of ultrasound images that has the highest quality was produced.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing images using an image processing neural network system. One of the systems includes a domain transformation neural network implemented by one or more computers, wherein the domain transformation neural network is configured to: receive an input image from a source domain; and process a network input comprising the input image from the source domain to generate a transformed image that is a transformation of the input image from the source domain to a target domain that is different from the source domain.

Abstract: Aspects of the technology described herein relate to collection and display of ultrasound images using an explaining model. A first ultrasound image may be determined to be in a first class and a second ultrasound image that is in a second class may be generated based on the first ultrasound image. The second ultrasound image may be generated by an explaining model. A classification model may classify the first and second ultrasound images in the first and second classes, respectively. Generating the second ultrasound image may include changing one or more portions of the first ultrasound image. The explaining model may also generate a transformed version of the first ultrasound image and a mask image, and the second ultrasound image may be a composite image of the first ultrasound image and the transformed version of the first ultrasound image. The mask image may determine how to generate the composite image.

Abstract: Aspects of the technology described herein relate to methods and apparatuses for identifying gestures based on ultrasound data. Performing gesture recognition may include obtaining, with a wearable device, ultrasound data corresponding to an anatomical gesture; and identifying the anatomical gesture based on the obtained ultrasound data. Interfacing with a computing device may include identifying, with a wearable device, an anatomical gesture using ultrasound data obtained by the wearable device; and causing the computing device to perform a specific function based on the anatomical gesture identified by the wearable device.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing images using an image processing neural network system. One of the system includes a shared encoder neural network implemented by one or more computers, wherein the shared encoder neural network is configured to: receive an input image from a target domain; and process the input image to generate a shared feature representation of features of the input image that are shared between images from the target domain and images from a source domain different from the target domain; and a classifier neural network implemented by the one or more computers, wherein the classifier neural network is configured to: receive the shared feature representation; and process the shared feature representation to generate a network output for the input image that characterizes the input image.

Abstract: Aspects of the technology described herein relate to configuring an ultrasound device with ultrasound imaging presets. Some embodiments include configuring an ultrasound device with different sets of imaging parameter values without receiving user input regarding configuration of the ultrasound device between configuration of the ultrasound device with one set of imaging parameter values and configuration of the ultrasound device with another set of imaging parameter values. Some embodiments include automatically configuring an ultrasound device with a set of imaging parameter values based on automatically determining an anatomical location being imaged by the ultrasound device. Some embodiments include automatically configuring an ultrasound device with a cardiac preset when a cardiac region is being imaged and automatically configuring the ultrasound device with an abdominal preset when an abdominal region is being imaged during a FAST or eFAST exam.

Abstract: A method of determining the position and orientation of an invasive medical device is described using altered devices and a machine learning algorithm for detecting the position and orientation of such devices from imagery. Such predictions can subsequently be displayed to an operator to improve the speed and accuracy by which they perform procedures.

Abstract: Surface reconstruction contour completion embodiments are described which provide dense reconstruction of a scene from images captured from one or more viewpoints. Both a room layout and the full extent of partially occluded objects in a room can be inferred using a Contour Completion Random Field model to augment a reconstruction volume. The augmented reconstruction volume can then be used by any surface reconstruction pipeline to show previously occluded objects and surfaces.

Abstract: Surface reconstruction contour completion embodiments are described which provide dense reconstruction of a scene from images captured from one or more viewpoints. Both a room layout and the full extent of partially occluded objects in a room can be inferred using a Contour Completion Random Field model to augment a reconstruction volume. The augmented reconstruction volume can then be used by any surface reconstruction pipeline to show previously occluded objects and surfaces.

Abstract: Surface reconstruction contour completion embodiments are described which provide dense reconstruction of a scene from images captured from one or more viewpoints. Both a room layout and the full extent of partially occluded objects in a room can be inferred using a Contour Completion Random Field model to augment a reconstruction volume. The augmented reconstruction volume can then be used by any surface reconstruction pipeline to show previously occluded objects and surfaces.