Back to results
US 12,731,365 B1Granted
Automatically reporting on an anatomical structure visualized in an ultrasound image that is selected by a user
AssigneeEchoNous, Inc.
InventorsBabajide Ayinde, Matthew Cook
Life of this patent
Estimating the term…
Abstract
A facility for presenting automatically generated information about an ultrasound image is described. The facility causes the ultrasound image to be displayed in a display region. Simultaneous to the display of the ultrasound image, the facility causes the following to be displayed in the display region: (1) quantitative data automatically determined for an anatomical feature of interest selected based on user input provided in connection with the display of the ultrasound image, and (2) qualitative data automatically determined for the anatomical feature of interest.
Claims20
A system, comprising:
- an ultrasound sensing device;
- and
- a computing device, the computing device comprising:
- a communication interface configured to directly receive ultrasound echo data sensed by the ultrasound sensing device from a patient, the received ultrasound echo data comprising an ultrasound image;
- a memory configured to:
- store one or more neural networks each trained to analyze ultrasound images;
- a processor configured to:
- cause the ultrasound image to be displayed;
- receive input originating from a user specifying one or more points within the displayed ultrasound image;
- subject to a first neural network inputs comprising the ultrasound image and the received input to obtain a bounding box within the ultrasound image inferred to surround an anatomical feature of interest;
- subject to a second neural network inputs comprising the ultrasound image and the obtained bounding box to obtain an identity of the anatomical feature of interest;
- subject to a third neural network inputs comprising the ultrasound image, the obtained bounding box, and the obtained feature identity to obtain a segmentation mask identifying pixels of the ultrasound image visualizing the anatomical feature of interest;
- subject to one or more first tools inputs comprising the ultrasound image, the obtained bounding box, the obtained feature identity, and the segmentation mask to obtain one or more quantitative attributes of the anatomical feature of interest;
- subject to one or more second tools inputs comprising the ultrasound image, the obtained bounding box, the obtained feature identity, and the segmentation mask to obtain one or more qualitative attributes of the anatomical feature of interest;
- and
- cause to be displayed simultaneously with the ultrasound image the obtained one or more quantitative attributes of the anatomical feature of interest and the obtained one or more qualitative attributes of the anatomical feature of interest.
13 more elements
A method in a computing system, comprising:
- receiving an ultrasound image;
- causing the ultrasound image to be displayed;
- receiving input originating from a user specifying one or more points within the displayed ultrasound image;
- subjecting to a first neural network inputs comprising the ultrasound image and the received input to obtain, with respect to an anatomical feature of interest selected by the received input, an identity of the anatomical feature of interest and a segmentation mask identifying pixels of the ultrasound image visualizing the anatomical feature of interest;
- subjecting to one or more first tools inputs comprising the ultrasound image, and the obtained feature identity to obtain one or more quantitative attributes of the anatomical feature of interest;
- and
- subjecting to one or more second tools inputs comprising the ultrasound image and the obtained feature identity, and the segmentation mask to obtain one or more qualitative attributes of the anatomical feature of interest.
4 more elements
One or more hardware memories collectively having contents configured to cause a computer system to perform a method, the method comprising:
- accessing an ultrasound image;
- causing the ultrasound image to be displayed;
- receiving input originating from a user specifying one or more points within the displayed ultrasound image;
- subjecting to a first neural network inputs comprising the ultrasound image and the received input to obtain, with respect to an anatomical feature of interest selected by the received input, an identity of the anatomical feature of interest and a segmentation mask identifying pixels of the ultrasound image visualizing the anatomical feature of interest;
- subjecting to one or more first tools inputs comprising the ultrasound image, and the obtained feature identity to obtain one or more quantitative attributes of the anatomical feature of interest;
- and
- subjecting to one or more second tools inputs comprising the ultrasound image and the obtained feature identity to obtain one or more qualitative attributes of the anatomical feature of interest.
4 more elements
A system, comprising:
- an ultrasound sensing device;
- and
- a computing device, the computing device comprising:
- a communication interface configured to directly receive ultrasound echo data sensed by the ultrasound sensing device from a patient, the received ultrasound echo data comprising an ultrasound image;
- a memory configured to:
- store one or more neural networks each trained to analyze ultrasound images;
- a processor configured to:
- cause the ultrasound image to be displayed;
- receive input originating from a user specifying one or more points within the displayed ultrasound image;
- subject to a first neural network inputs comprising the ultrasound image and the received input to obtain, with respect to an anatomical feature of interest selected by the received input, an identity of the anatomical feature of interest and a segmentation mask identifying pixels of the ultrasound image visualizing the anatomical feature of interest;
- subject to one or more first tools inputs comprising the ultrasound image and the obtained feature identity to obtain one or more quantitative attributes of the anatomical feature of interest;
- and
- subject to one or more second tools inputs comprising the ultrasound image and the obtained feature identity to obtain one or more qualitative attributes of the anatomical feature of interest.
10 more elements
Dependencies are inferred from the claim text.